Globalization and the Wage-Working Conditions Relationship ...
Transcript of Globalization and the Wage-Working Conditions Relationship ...
Macalester CollegeDigitalCommons@Macalester College
Honors Projects Economics Department
5-1-2009
Globalization and the Wage-Working ConditionsRelationship: A Case Study of CambodianGarment FactoriesCael WarrenMacalester College, [email protected]
Follow this and additional works at: http://digitalcommons.macalester.edu/economics_honors_projects
Part of the International Economics Commons, and the Labor Economics Commons
This Honors Project is brought to you for free and open access by the Economics Department at DigitalCommons@Macalester College. It has beenaccepted for inclusion in Honors Projects by an authorized administrator of DigitalCommons@Macalester College. For more information, pleasecontact [email protected].
Recommended CitationWarren, Cael, "Globalization and the Wage-Working Conditions Relationship: A Case Study of Cambodian Garment Factories"(2009). Honors Projects. Paper 25.http://digitalcommons.macalester.edu/economics_honors_projects/25
1
Globalization and the Wage-Working Conditions Relationship:
A Case Study of Cambodian Garment Factories
Cael Warren
May 2009
Abstract
The wage premiums for firm-level foreign exposure (exporting and foreign ownership)
have been well documented in the literature, and their potential sources have been studied in
depth. Compensating differentials and efficiency wages are two distinct explanations (with
radically different implications for worker welfare) for wage gaps that persist between firms
despite controls for firm and worker characteristics. We use a comprehensive dataset of working
conditions and wage compliance in Cambodia’s exporting garment factories to explore (1) the
impact of foreign ownership on wages and working conditions, (2) whether the relationship
between wages and working conditions within these exporting factories more closely resembles
efficiency wage or compensating differential theory and (3) whether the wage-working
conditions relationship differs between domestically owned and foreign-owned firms.
We find that foreign ownership increases compliance on both wages and working
conditions, contradicting the contention that higher wages in foreign-owned firms compensate
workers for worse working conditions. In addition, we find a robust positive relationship
between wages and working conditions in the sample as a whole, suggesting that efficiency
wages or a similar theory more accurately explains the behavior of these exporting firms than
compensating differentials. This positive relationship is stronger in domestically owned firms
than in foreign-owned firms, but the relationship remains positive, fairly large, and statistically
significant even in foreign-owned firms. Due to the lack of evidence in support of compensating
differential theory, we conclude that both foreign ownership and exogenously imposed
improvements in working conditions improve net worker welfare.
2
Despite the conventional wisdom that foreign-owned factories in developing countries
operate as “sweatshops,” paying low wages and providing unpleasant work environments, many
studies have shown that wages are higher in foreign-owned firms than in their otherwise identical
domestically owned counterparts.1 The literature has also shown that exporting firms pay higher
wages than non-exporting firms,2 lending further support to the notion that working in a
“globalized” firm benefits workers. These results are encouraging, but they do not necessarily
imply that exposure to foreign markets improves worker welfare overall. If higher wages
compensate workers for poor working conditions, workers may be no better off in these firms.
If, on the other hand, wages do not decline as working conditions improve, workers may be
made better off by working in a foreign-owned or exporting firm. Determining the presence (or
absence) of compensating differential relationships in exporting and foreign-owned firms is thus
critical to understanding the impact of globalization on workers in developing countries.
The literature consistently reveals positive wage premiums in exporting and foreign-
owned firms relative to non-exporting and domestically owned firms, but the source of these
wage premiums remains unclear. Using a detailed dataset of exporting factories in Cambodia,
this paper explores (1) how wages and working conditions differ between domestically and
foreign-owned firms, (2) whether compensating differentials explain the wage changes that occur
within the full sample of domestically and foreign-owned exporting firms over time and (3)
whether the relationship between wages and working conditions differs between domestically
and foreign-owned firms. We find that foreign-owned firms are more compliant than
domestically owned firms on both wages and working conditions, suggesting that compensating
differentials cannot explain the foreign ownership wage premium in these factories. In addition,
1 See, for example, Aitken et al. (1996), Girma and Görg (2007), and Sjöholm and Lipsey (2006).
2 Bernard and Jensen (1995), Glick and Roubaud (2006), and Schank et al. (2007) are a few examples.
3
good working conditions are positively related to wages within firms, suggesting that
improvements in working conditions do not induce firms to reduce wages. This positive
relationship is stronger in domestically owned firms, but is also positive, relatively large, and
statistically significant in foreign-owned firms. Due to this evidence contradicting compensating
differential theory both between domestically and foreign-owned firms and within firms, we
move one step closer to the conclusion that both foreign ownership and improvements in
working conditions make workers in these factories better off overall.
Firms exposed to foreign markets tend to pay higher wages, even when controlling for a
variety of factors. Several studies, in both developing and developed countries, have shown that
foreign-owned firms pay higher wages than their domestically owned counterparts, controlling
for many firm and worker characteristics. Aitken et al. (1996), Girma and Görg (2007), and
Sjöholm and Lipsey (2006), are just a few examples of such studies.3 Exporting firms also tend
to pay higher wages than non-exporting firms, controlling for a variety of firm characteristics.
Several studies have verified this trend in a variety of contexts, from manufacturing plants in the
U.S. (Bernard and Jensen 1995) 4 to Export Processing Zones in Madagascar (Glick and
Roubaud 2006) to exporting firms in Germany (Schank et al. 2007). Though a few other studies
have failed to show evidence of this relationship, the preponderance of the evidence seems to
suggest that exporting firms pay higher wages than non-exporting firms.
Compensating differentials and efficiency wages, two theories with opposite implications
for worker welfare, are the literature’s dominant explanations for wage gaps that persist between
firms despite controls for firm characteristics. The evidence supporting the efficiency/fair wage
3 See Brown et al. (2002) or Lipsey (2004) for a more comprehensive review of the literature on the ownership-
wage relationship. 4 The wage premium in exporting firms persists despite a variety of controls and plant-level fixed effects. Though
the bulk of the premium is explained by other firm-level controls like plant size, capital intensity, hours per worker,
industry, and location, the premium for exporting firms remains.
4
model is extensive, indicating that firms often pay above-market wages to harness productivity
gains. Empirical evidence has shown that paying efficiency wages reduces shirking (Cappelli
and Chauvin 1991), increases worker effort (Goldsmith et al. 2000), increases worker
productivity (Fuess and Millea 2002), and increases the firm’s market share through those
productivity gains (Konings and Walsh 1994). Arai (1994) finds indirect evidence that firms are
using higher wages to reduce shirking, showing that Swedish inter-industry wage differentials
are strongly and positively related to levels of worker autonomy. The literature thus suggests the
presence of efficiency wage behavior among firms, a sign that higher wages could signal a net
improvement in welfare for the workers receiving them (since the higher wages yield output
increases for the firm, thereby eliminating the need for cost-cutting working conditions
reductions in response to the wage increases).
Empirical tests of compensating differential theory, meanwhile, have turned up mixed
results. While many have found evidence of compensating differentials for accident risk
(Cousineau et al. 1992; Marin and Psacharopoulos 1982), occupation- and industry-level work-
related mortality risk (Leigh 1991), hard, physical, or stressful work (Duncan and Holmlund
1983; Duncan and Stafford 2002 [1980]) and inconvenient work hours (Duncan and Holmlund
1983; McNabb 1989; Altonji and Paxson 1988), others have found little evidence of
compensating differentials for these working conditions and others (Brown 1980; Dorman and
Hagstrom 1998; McCrate 2005).5 In addition to its inconsistent support for compensating
differential theory, the literature is also entirely comprised of worker-level studies despite the
firm’s essential role in determining wages and working conditions. The mixed results in the
literature may be due in part to this lack of firm-level studies. Nonetheless, the results suggest
5 A few studies apply compensating differential theory to industry-level export wage premiums (using worker-level
data), and they too find little or no evidence of compensating differentials in El Salvador (Robertson and Trigueros-
Argüello 2008), Indonesia (Robertson et al. 2008), and Cambodia (Robertson and Neak 2008).
5
that workers might gain a net increase in welfare from higher wages, but the higher wages
sometimes compensate them for otherwise worse working conditions.
The scarcity of firm-level working conditions data has so far meant that studies of the
firm’s choice between employing efficiency wages or compensating differentials in worker
compensation are very rare. Furthermore, the minimal diversity of working conditions measures
available in most datasets, even at the worker level, has prevented a close examination of the full
package of wages and working conditions offered. Finally, while many have compared wages in
domestically and foreign-owned firms, none have studied whether the higher wages in foreign-
owned firms are connected to worse working conditions. This paper, using a comprehensive
dataset of working conditions in Cambodia’s exporting garment factories from the Better
Factories Cambodia (BFC) program, explores this wage-working conditions relationship to
evaluate the net impact of foreign ownership and working conditions improvements on worker
welfare.
The influence of the Better Factories Cambodia program (described in section three) in
these firms provides a unique situation with great empirical potential. While most firm-level
studies must rely on various immeasurable or random exogenous shocks for their data variation,
BFC provides a common and known shock across firms, applying pressure on all firms to
improve working conditions and wage compliance.6 With this great empirical strength of the
dataset, however, come two limitations of note. First of all, the dataset contains only measures
of wage compliance, not of worker compensation itself. We therefore use an index of five
measures of wage compliance (explained in detail in section three) to proxy for wages.
Secondly, because the dataset is entirely comprised of exporting firms, we cannot explore both
6 This is not to say that the BFC effect is uniform across firms, but we account for the heterogeneity of the BFC
effect with firm-level controls for the cumulative number of BFC visits and their frequency.
6
the exporting and ownership dimensions of the effect of foreign exposure on the wage-working
conditions relationship.
Instead, we explore the impact of foreign ownership within this sample of exporting firms
in three steps. First (in section four), we identify the positive effect of foreign ownership on both
wages and working conditions in these firms, controlling for observable firm characteristics.
This finding contributes to the limited literature on the foreign ownership wage premium, but
says nothing about how the firms choose combinations of wages and working conditions over
time, particularly in response to a shock. We therefore examine the wage-working conditions
relationship within firms over time, revealing the firm’s choice between the compensating
differential and efficiency wage approaches to worker compensation. In section five, we explore
this wage-working conditions relationship within the entire sample of foreign-exposed
(exporting) firms.7 In section six, we examine how that wage-working conditions relationship
differs by the dimension of foreign exposure for which our dataset contains variation –
ownership. Before we proceed with the empirical results, however, we will lay out a firm-level
theoretical framework to illustrate the contrasting predictions of the compensating differential
and efficiency wage models, and then describe the dataset used to evaluate these theoretical
predictions.
2. Conceptual Framework
To compare the predictions of the wage-working conditions relationship presented by the
theories of compensating differentials and efficiency wages, we apply a basic isoquant
production framework that is based on five assumptions. First, firms respond rationally to an
7 This component says nothing about the effect of foreign exposure, but examines firm behavior in selecting
combinations of wages and working conditions within a unique dataset of foreign-exposed firms.
7
exogenously imposed positive shock to working conditions.8 Second, firms differentiate
themselves according to output quality, as demonstrated specifically in exporting firms by
Mandel (2008). Third, firms can improve output quality by eliciting more effort from workers.
Fourth, workers will put forth more effort if they receive greater compensation, which is
comprised of combinations of wages and working conditions. Finally, workers are willing to
trade off wages and working conditions as inputs in their “production” of effort for the firm.
There are many combinations of wages and working conditions that a firm can offer to
elicit each intended level of quality/effort from workers. Because workers trade off wages and
working conditions in their effort production function, wages and working conditions are
negatively related within a given level of quality/effort. A graphical depiction of the firm’s
problem is illustrated in Figure 2a. A firm aiming to elicit a low level of effort might operate
anywhere on the Low Effort isoquant. One such firm, starting at some combination of wages and
working conditions represented by point P, has two broad options for the path it takes in
wages/working conditions space when an exogenous improvement in working conditions is
imposed. It can reduce wages in response to the higher costs of improving working conditions
(move down along the Low Effort isoquant to point N) or it can hold wages constant or even
increase them (move to the High Effort isoquant, to point M). Moving along a given effort curve
represents the wage-working conditions tradeoff, or the compensating differential relationship.
A shift to a higher effort curve, meanwhile, illustrates the essence of efficiency wage theory:
8 This theoretical analysis considers an exogenous improvement in working conditions as prompted by the Better
Factories Cambodia program. The conclusions would be the same if we considered an exogenous improvement in
wages and its impact on working conditions, because this analysis considers the relationship between wage
compliance and working conditions, not the causality therein. In other words, for each improvement in either wages
or working conditions made by the firm, this model considers the two possible effects (negative or nonnegative) on
the other form of compensation.
8
increasing total worker compensation can be profit-maximizing for the firm when it produces
greater worker effort.
As a result, the relationship between wages and working conditions within firms over
time can reveal their choice between a compensating differentials approach (which holds worker
welfare constant despite changing compensation mixes) and alternative approaches such as
efficiency wages (which improve worker welfare). The next section describes the data that we
will use to explore this wage-working conditions relationship in Cambodian garment factories.
3. Data
In this section, we detail the data that we use to empirically examine the relationships
between foreign ownership, wage compliance, and working conditions compliance. First, we
describe the source of the dataset, its contents, and the design of the program that supplied it.
Next, we describe how we combine the numerous working conditions and wage compliance
measures into a few comprehensive indicator variables for empirical analysis. Finally, we
provide summary statistics of the variables we use.
3.1 Data Source
The data come from the Better Factories Cambodia (BFC) program of the International
Labor Organization. Designed to improve working conditions in Cambodian factories by
addressing the problem of imperfect information between factories and buyers, this program
aims to inform buyers about the conditions in the factories from which they purchase garments.
To do so, BFC monitors working conditions in all Cambodian garment factories during
unannounced visits, sending Cambodian monitors into factories to complete a survey assessing
the factory’s compliance on a variety of working conditions and wage requirements. To avoid
monitor bias, each monitoring team contains at least two people, and the same team rarely
9
assesses the same factory twice. After the factory’s second BFC visit, BFC publishes the firm’s
name and progress on improving working conditions in an annual synthesis report, which they
share with the factories’ buyers.
As the Cambodian government has mandated that all exporting garment factories consent
to this monitoring program, it eventually reached all such factories. The original wave of visits
in 2001-2002 reached 119 factories with the first survey created for BFC. For the three years
following the visits to these original factories, monitors conducted visits using less formal
techniques and did not carefully record results, so data are unavailable for this three-year period.
The next wave of documented visits began with the launch of the improved Information
Management System (IMS) survey in December 2005. Since then, monitors have visited each
factory an average of once every eight months. Through July 2008, this panel dataset contains
363 factories and 1154 factory-visit observations, of which 289 factories have more than one
visit and a known country of origin (for a total of 1060 observations).
The theoretical framework calls for variables representing wages, working conditions,
and the standard determinants of wages within firms such as size, age, and ownership (Brown
and Medoff 1989; Brown and Medoff 2003). Because wages themselves are unavailable in the
dataset, an index of five measures of compliance on wage law (explained below) will serve as a
proxy for wages. The dataset contains approximately 130 measures of working conditions,
which we aggregate in different ways to represent working conditions empirically. Firm controls
include firm age (in months), firm size (measured as the total number of workers) and the
percentage of workers in a union, all of which should predict higher wage compliance. We also
control for the variation in the BFC effect using measures of the cumulative number of BFC
10
visits and their frequency. Finally, specifications in section six will include a control for foreign
ownership.
3.2 Construction of Index Variables
The dataset includes approximately 130 compliance variables, all on a 0/1 compliance/
noncompliance scale. The compliance questions from which these variables originate, matched
between the original and IMS surveys, are listed in Appendix A. To make these useful for
analysis, we group these variables into four broad working conditions categories (shown in Table
3a) with several subcategories within each category. We generate compliance rates for each
category as the simple average of compliance across the questions in the category, normalized to
a scale of 100. Wages, for example, contains five compliance questions9, so a Wages value of 60
means that the factory was compliant on three of the five wage payment questions during that
visit. We generate all other indices in the same way, though the rest contain more questions,
ranging from 13 to 43 in the disaggregated working conditions measures. The most complicated
index is Working Conditions, which contains all of the other non-wage indices shown in Table
3a, and is the measure of working conditions used in this paper unless specified otherwise.
3.3 Summary Statistics
The working conditions covered by the survey range from occupational safety and health
(OSH) to freedom of association and collective bargaining (FACB) to maternity leave and other
benefits. The categories of working conditions and the summary statistics of their compliance
rates, along with some basic firm characteristics and the breakdown of ownership groups, are
shown in Table 3b. The average factory is almost five years old and employs about 1200
workers. Of the 363 factories, 278 have received at least two BFC visits and have complete data
9 The five compliance variables included in the Wages index are whether the firm paid the proper minimum wage,
overtime wage, night wage, holiday wage, and wage during weekly time off (Sunday).
11
for the necessary firm controls. Visits typically fall about ten months apart, but the time between
visits varies widely due to a gap in the dataset (explained below). As shown in Table 3c, the vast
majority of the sample (95%) is foreign-owned, with about 65% owned by Taiwan, Hong Kong,
and China; 22% owned by Korea, Malaysia, and Singapore; 3% owned by Western countries;
and 2% owned by other Asian countries.
The mean level of working conditions compliance in the sample was about 86%, meaning
that the average factory visited between 2001 and 2008 was found to be noncompliant on about
14% of measures. The mean level of wage compliance is higher (92%), but it also varies more
widely. Rates of compliance on the smaller working conditions categories range from the
relatively low 81% on OSH to the relatively high 91% for FACB.
Finally, Table 3d illustrates the varying levels and changes of wage and working
conditions compliance by different ownership groups and in different periods. In general,
compliance is fairly high and improving for most groups, with the exception of wage compliance
in Cambodian firms. Malaysian firms tended to be the most compliant on both wages and
working conditions, while Cambodian firms were the least compliant on these measures.
Chinese firms improved working conditions at the fastest rate, while Other Asian firms improved
wages at the fastest rate. Most interestingly, foreign-owned firms exhibited greater compliance
on both wages and working conditions as well as greater improvement in compliance on wages
than domestically owned firms. These statistics give no indication of a compensating differential
relationship between wages and working conditions, as the groups most compliant on wages are
also the most compliant on working conditions. We turn next to statistical analysis to further
explore this question.
12
4. Foreign Ownership’s Impact on Wages and Working Conditions
For a basic idea of one aspect of globalization’s effect on the welfare of workers in these
garment factories, we first explore the impact of foreign ownership on wages and working
conditions. We begin by estimating Equation (1), shown below:
Wagesit = ß0 + ß1(FirmSizeit)+ ß2(FirmAgeit) + ß3(%Unionit) + ß4(ForeignOwnershipit) + εit (1)
where t is measured in visits, i is the firm, Wages is an index variable as described above, Firm
Size is the number of workers employed by the firm, Firm Age is measured in months, %Union
is the percentage of workers in a union, and Foreign Ownership is a dummy variable equal to
one if the firm is not Cambodian-owned. The results, shown in the first column of Table 4a,
indicate a relatively large and statistically significant (at the 10% level) effect of Foreign
Ownership on wage compliance, with wage compliance about nine percentage points higher in
foreign-owned firms than in domestically owned firms. These results confirm findings
elsewhere in the literature of higher wages in foreign-owned firms, so long as we assume wage
compliance to be an effective proxy for wages.
These results might be biased by the fact that firms have differing numbers of
observations. If there is a systematic relationship between a factory’s number of visits, its
ownership status, and its wage compliance, including multiple visit observations for each firm
could bias our results in some way. We therefore run a regression between firms, essentially
evening out the number of observations per firm. The result of this change, shown in column
two of Table 4a, is very little change in the magnitude of the foreign ownership coefficient and a
small increase in its statistical significance (which can be explained by the fact that the standard
errors for between regressions cannot be corrected for heteroskedasticity). Our results therefore
13
appear not to be biased by varying number of observations per firm, suggesting that foreign
ownership does indeed have a positive effect on wage compliance.
The positive effect of foreign ownership on wage compliance does not, however,
guarantee that workers in foreign-owned firms are better off than those in domestically owned
firms. We therefore also examine the effect of foreign ownership on the index of working
conditions, running Equation (1) with Working Conditions (the aggregated index as described
above) as the dependent variable. The third column of Table 4a presents the results, which show
a strong and statistically significant effect of foreign ownership on working conditions
compliance. While foreign ownership has a smaller effect on working conditions (about a four-
percentage-point increase) than on wages, the coefficient is still fairly large and statistically
significant at the 1% level. When we look at the foreign ownership on working conditions in a
between-firms regression, the magnitude and significance of the coefficient both fall slightly, but
the positive and statistically significant sign remains. Since foreign ownership appears to have a
strong and statistically significant impact on both wages and working conditions, these results
suggest that higher wages (represented by greater wage compliance) in foreign-owned firms do
not serve as compensating differentials for worse working conditions.
Because the detailed nature of our dataset allows us to explore further details of the
foreign ownership relationship with wage compliance and working conditions, we disaggregate
the foreign ownership variable into the eight countries/groups of countries shown in Table 3c and
include indicator variables for each in place of the foreign ownership dummy in Eq. (1). The
results, shown in column one of Table 4b, reveal that the bulk of the foreign ownership
coefficient results from the large and statistically significant positive coefficients on Korea,
Malaysia, and Singapore. Interestingly, when we run the between regression (column two of
14
Table 4b), we find that Hong Kong also carries a large and statistically significant coefficient,
though the results change very little otherwise. Clearly, the effect of foreign ownership on wage
compliance is not universally identical; the source of the foreign ownership determines the
magnitude and significance of its effect.
The same is true of the positive effect of foreign ownership on working conditions. The
results of the random effects regression, with Working Conditions as the dependent variable,
reveal positive and statistically significant effects of all countries/groups but China and Other
Asia. Looking at the between effects results (column four of Table 4b), we see that West and
Singapore lose their statistical significance, and the significant country coefficients again fall in
magnitude, but the positive and statistically significant effect remains. These results confirm that
the country of origin impacts the magnitude and significance of the foreign ownership effect.
While the specific country of ownership matters, disaggregating the foreign ownership variable
does allow us to see that the positive Foreign Ownership coefficient is no fluke; foreign
ownership does appear to improve working conditions and wage compliance relative to
Cambodian ownership.
5. Wages and Working Conditions Within Firms Over Time
The positive effect of foreign ownership on wages and working conditions separately
says little about how firms choose combinations of wages and working conditions, but this firm
choice is vital to workers’ welfare outcomes. Understanding the impact of changing working
conditions on wage compliance within firms, especially in response to an exogenous shock like
the implementation of Better Factories Cambodia, can help reveal whether such programs have a
net positive impact on workers. We therefore now consider the relationship between wages and
working conditions within firms over time in the full sample of exporting garment factories.
15
5.1 Estimation Issues
While the small number of time periods mitigates the risk of serial correlation or
nonstationarity, the wide diversity of the firms makes heteroskedasticity likely. Results of a
Breusch-Pagan/Cook-Weisberg test confirm this suspicion. The empirical results that follow
report heteroskedasticity-corrected standard errors to address this issue. In addition,
multicollinearity could be a concern. Diagnostic analysis suggests only mild multicollinearity, 10
so we will proceed, acknowledging that there are some moderate correlations between
explanatory variables, especially when we disaggregate working conditions.
Finally, the potentially simultaneous determination of wages and working conditions
means that OLS estimation could yield biased coefficients in a standard statistical analysis, since
the simultaneity leads to a correlation between the Working Conditions variable and the error
term. In a typical analysis aiming to assess a causal relationship between a dependent and
independent variable, this simultaneity would bias the regression results and undermine their
validity. In our case, however, we aim to make no statements about the causal relationship
between working conditions and wage compliance. We instead aim to analyze the firms’
simultaneous decisions of wage-working conditions combinations. Whether wage compliance
affects working conditions or vice versa, the sign of the coefficient tells us whether firms
improve or worsen their compliance on one when they improve on the other. It is the sign of this
relationship, no matter the direction of the causal arrow, in which we are interested. Because our
interpretation of the coefficients differs in this way from the typical analysis, our conclusions are
not biased by the simultaneous determination of working conditions and wage compliance.
5.2 Initial Results
10
Among the simple correlation coefficients between categories, no coefficient exceeds 0.6, though one exceeds
0.5. The remainder of the correlation coefficients are less than 0.25. A test of the Variance Inflation Factors
indicates only mild multicollinearity, with a maximum VIF of 1.6.
16
The compensating differential literature guides us with two analytical techniques for
evaluating the wage-working conditions relationship. The first method we explore includes
dependent and independent variables in the current period, with fixed effects to absorb any firm-
based variations in productivity or other omitted controls. We begin by estimating Equation (2)
below, where t is measured in visits, i is the factory, Wages and Working Conditions are indices
as described above, Firm Size is in hundreds of workers, Firm Age is in years, %Union is the
percentage of workers in a union, Visit is the number of visits completed (including the t’th
visit), and Time is the number of months since the last BFC visit to the factory.
Wagesit = ß0 + ß1(Working Conditionsit) + ß2(Firm Sizeit) + ß3(Firm Ageit) + ß4(%Unionit) +
ß5(Visitit) + ß6(Timeit) + εit (2)
The results, shown in column one of Table 5a, are a surprising contradiction to
compensating differential theory but correspond well with the results of Section 4. While none
of the controls is statistically significant, most are correctly signed, and the Working Conditions
coefficient is positive, relatively large, and statistically significant at the one percent level. The
coefficient of 0.783 indicates that, for each ten percent improvement in working conditions
compliance, wage compliance increases almost eight percent. This pattern emerges despite our
controls for the firm age, firm size, unionization in the firm, number of BFC visits to the factory,
and amount of time since the last BFC visit. Explanatory power of the regression is low,
however, with an overall R-squared of only 0.08, and the controls are all statistically
insignificant when we use heteroskedasticity-corrected standard errors. Nonetheless, these
results indicate that, controlling for the theoretically essential firm characteristics, working
conditions and wage compliance are positively related.
17
These results, however, fail to capture the main advantage of the fixed effects method
relative to the difference-in-difference method; using fixed effects allows us to consider a larger
sample size because we can include the first visit in the time series. In this particular
specification, however, the Time variable is measured as the time between visits, thereby
excluding the first observation for each firm from the regression. Given the statistical
insignificance of the Time control, its exclusion seems warranted to enable a broader
examination of the relationship. Excluding this variable, the results of which are shown in
column two of Table 5a, increases the sample size by over fifty percent. The results are quite
similar to those of column one, with a slight increase in the magnitude of the coefficient but no
change in its significance. These results indicate a strong and relatively large positive
relationship between wages and working conditions in these firms, regardless of whether we use
a specification that captures the full sample.
The other analytical method most frequently used to identify compensating differentials
is the difference-in-difference approach. Because this method has generally been more effective
in identifying compensating differential relationships, and because the two levels regressions
suggest no major change in results when using the larger sample size, the rest of our analysis will
employ the difference-in-difference approach. 11
This regression equation, shown below,
explores the relationship between the change in wage compliance and the change in working
conditions compliance.
∆Wagesi(t-[t-1]) = ß0 + ß1(∆Working Conditions i(t-[t-1])) + ß2(∆Firm Size i(t-[t-1])) + ß3(Firm Ageit) +
ß4(∆%Union i(t-[t-1])) + ß5(Visitit) + ß6(∆Time i(t-[t-1])) + εit (2a)
11
The difference-in-difference approach allows us to examine changes within firms over time, holding constant any
firm-specific variation unobserved in other control variables. This approach is commonly used in the compensating
differential literature to control for productivity variation among units of observation (in our case the firm; in most
cases the worker), and appears to be the only empirical method to consistently illustrate the theoretically predicted
compensating differential relationship.
18
Regression results for Equation (2a), shown in the third column of Table 5a, illustrate a
fairly strong positive relationship between working conditions and wage compliance in these
firms. The statistically significant coefficient of 0.869 indicates that, when the change in
working conditions compliance improves by ten percentage points, the change in wage
compliance improves by nearly nine percentage points.12
In other words, improving working
conditions translates almost one-for-one into improving wage compliance.
These results contradict the contention of compensating differential theory that wages
and working conditions should move opposite one another within firms. The observed positive
relationship between working conditions and wage compliance implies that these firms can
improve their outcomes by increasing their total compensation mix to workers (moving from the
Low Effort to the High Effort isoquant); if this were not the case, the firm’s rational behavior
would lead to a negative relationship between wages and working conditions. It appears,
therefore, that the efficiency wage model, which predicts simultaneous improvements in wages
and working conditions (presumably) to inspire greater worker effort, captures the behavior of
these exporting firms better than the compensating differentials model. While we cannot
contrast these results with those of non-exporting firms, we can say that, within this sample of
foreign-exposed firms, higher wage compliance does not signal worse working conditions or
vice versa.
5.3 Robustness
To evaluate the robustness of the large and significant working conditions coefficient, we
use a variety of alternative specifications and sample alterations, the results of which we will
discuss in this subsection. First of all, given the subjective nature of the data collection and the
12
Recall that both wages and working conditions are measured in indices of compliance, generated in such a way
that a one-unit increase amounts to a one percentage point improvement in compliance.
19
discrete (0/1) nature of the compliance measures, the data could contain monitor-based variation
as different monitors draw different lines between compliance and noncompliance. We therefore
include a set of monitor dummy variables, equal to one if the monitor was present in the factory
for that visit. The results of including this set of dummy variables are shown in column four of
Table 5a. The dummy variables’ coefficients (not shown) are all statistically insignificant, and
the main effect of their inclusion is to increase the magnitude of the (still statistically
insignificant) Visit variable. The coefficient on Working Conditions increases slightly, and
remains statistically significant at the 1% level. The variation in monitors in the sample appears
not to affect the strong wages-working conditions relationship.
While unionization is a theoretically essential determinant of wage compliance, the data
used to generate the unionization variable are imperfect, and including this variable reduces the
sample by 160 observations. We therefore test whether these data imperfections or sample
limitations are somehow driving the strong relationship between wage compliance and working
conditions. Column five of Table 5a shows the results of Equation (2a) with unionization
excluded. The Working Conditions coefficient falls slightly, to 0.802, in response to this change,
but remains relatively large and statistically significant at the 1% level. Excluding each of the
other firm-level controls individually (not shown) has even less of an effect on the Working
Conditions coefficient and the other coefficients in the regression.13
It is also possible that wage compliance and working conditions move together simply
because both have improved over time, due to increasing standards globally and especially due
to the effect of BFC’s presence. Though we control for the variation in the BFC effect using the
number of visits and the time since the last visit, the global improvement over time may only be
13
Excluding Firm Size had the largest effect among these, reducing the Working Conditions coefficient to 0.85 (still
statistically significant at 1%) and having almost no effect on the other coefficients.
20
captured in a continuous time variable. We therefore include Time in the next specification, the
results of which are shown in column six of Table 5a. The coefficient on the Time variable is
positive but statistically insignificant, and its inclusion actually slightly increases the Working
Conditions coefficient. Wage compliance and working conditions may be improving together
over time, but taking out the time effect does not reduce the strength of the wage-working
conditions relationship.
Given the large gap in the dataset (explained briefly in section 3.1), we suspect that there
may be differences between the firms present in the first wave of visits in 2001-2002 and the
firms that entered the program when the new “IMS” system was launched in late 2005. Columns
one and two of Table 5b therefore estimate Equation (2a) separately for these two groups of
firms. While the Working Conditions coefficient remains virtually unchanged, these two columns
reveal some interesting differences between these two groups of firms. The effect of the amount
of time between visits is zero in the original firms, but negative and statistically significant (as
expected) among the IMS firms.14
The number of visits has the expected positive effect among
the original firms, but its coefficient is relatively large, negative, and statistically significant for
the IMS firms. 15
Surprisingly, given these other differences between the two groups, the
Working Conditions coefficient is almost the same for each sample as for the sample as a whole.
Combining these two groups appears not to mask any hidden negative relationship between wage
compliance and working conditions.
14
This difference is likely driven by the large gap in the dataset, which affects the time between visits one and two
for the original firms but not for the IMS firms. 15
This contrast suggests a potentially nonlinear relationship between visits and wage compliance over time, since
the original factories are earlier in the sample, but adding a visits-squared term (results not shown) yielded
statistically insignificant coefficients on the Visit variables and had no effect on the Working Conditions coefficient.
It seems that, despite the differences between these two groups of factories, the specification for the sample as a
whole does not improve with changes to the way the Visit variable is specified. We also generated a dummy
variable equal to one if the factory was one of the original factories, included that in the whole-sample regression,
and also included that dummy interacted with the Visit variable. The Working Conditions coefficient was
unaffected, and the other variables’ coefficients were statistically insignificant.
21
Examining the full sample could also mask differing cultures of compliance in more
compliant firms, leading to differing wage-working conditions relationships. In other words,
some firms, possibly those under certain ownership or with greater exposure to working
conditions enforcement officials, might simply be more compliant as a whole, thereby biasing
our results in favor of a stronger positive wage-working conditions relationship. We therefore
split the sample, roughly in half, by each firm’s average level of compliance over its lifetime in
the sample. Results of Equation (2a) for the more compliant firms (greater than 85% average
compliance over all of the firm’s visits for all compliance points, both wages and working
conditions) are shown in column three of Table 5b. Interestingly, the results are opposite what
we expected; while a culture of compliance would lead to a larger positive relationship in more
compliant firms, we observe a smaller positive relationship in higher-compliance firms. This
result may be attributable to the closed nature of the compliance score (the fact that maximum
compliance of 100% is attainable). Since 86% of the high-compliance firms have reached 100%
wage compliance, improvements in working conditions compliance in these firms can be
associated at best with no change in wage compliance, leading to a smaller (but still positive and
statistically significant) relationship between wages and working conditions in these firms, with
a coefficient magnitude about half as large as in the entire sample.
Isolating the lower-compliance firms, meanwhile, allows us to observe the larger positive
wage-working conditions relationship in these factories. The size of the firm and the degree of
unionization also become statistically significant positive predictors of greater wage compliance
in these lower-compliance firms. The contrasting wage-working conditions relationships
between high- and low-compliance factories is robust to the compliance percentage at which we
22
split the sample, consistently yielding a Working Conditions coefficient of around 0.4 for high-
compliance firms and 1.0 - 1.4 for low-compliance firms.16
It is also illustrative to split the sample by the compliance level of the observation rather
than averaged over the life of the firm. As shown in the previous set of results, when we divide
the sample by the firms’ average level of compliance over their lifetimes, more compliant firms
tend to exhibit smaller positive wage-working conditions relationships. If this is indeed due to
their inability to improve wages beyond 100% compliance, we should observe the same pattern
when we sort the sample by overall compliance in each firm-visit observation and divide the
sample according to this measure. Interestingly, while 93% of the high-compliance
observations17
in this sample have reached 100% wage compliance, the high-compliance
sample’s Working Conditions coefficient is roughly the same magnitude as (and, in fact, slightly
larger than) that of the low-compliance sample. These results, shown in columns five and six of
Table 5b, also contain similarly insignificant coefficients on control variables. Splitting the
sample by overall compliance at the observation level thus yields different results than when we
split by compliance at the firm level, but no sample exhibits the expected negative wage-working
conditions relationship that compensating differential theory predicts. We’ve therefore presented
some food for thought, but have yet to find any evidence supporting compensating differential
theory.
5.4 Disaggregated Working Conditions
The aggregated Working Conditions variable, generated as an index of 130 different
individual measures of working conditions, conceals a lot of variation among different types of
working conditions. Another interesting test of the results’ robustness, therefore, is to
16
We split the sample at 83% and 87% average compliance to find these results. Splitting at higher or lower
averages resulted in samples too small to effectively interpret results. 17
(where high-compliance is greater than 87% overall compliance on wages and working conditions combined)
23
disaggregate the Working Conditions variable into four broad categories (those shown in Table
3a). Replacing the aggregated Working Conditions variable in Equation (2) with these four
disaggregated variables yields Equation (3) below, the results for which are shown in column
one of Table 5b.
∆Wages i(t-[t-1]) = ß0 + ß1(∆OSH i(t-[t-1])) + ß2(∆Paperwork i(t-[t-1])) + ß3(∆FACB i(t-[t-1])) +
ß4(∆Internal Relations/Benefits i(t-[t-1])) + ß5(∆Firm Size i(t-[t-1])) + ß6(Firm Ageit) +
ß7(∆%Union i(t-[t-1])) + ß8(Visitit) + ß9(∆Time i(t-[t-1])) + εit (3)
With the disaggregated working conditions variables, the control variables remain
generally insignificant and of the same signs as in the previous specifications, and explanatory
power remains low, with an R-squared value of 0.09. Three of the four working conditions
variables are statistically significant, two of them at the 1% level. Paperwork, the index of
worker information, documentation, and communication with the Cambodian Labor Ministry,
carries a relatively large and statistically significant coefficient, an unsurprising result given that
compliance improvements in this category are relatively low cost and therefore less likely to be
traded off with wage compliance. Controlling for the level of unionization, Freedom of
Association and Collective Bargaining (FACB) carries a positive coefficient that is significant
only at the 10% level. In other words, even when we control for the positive effect of
unionization on wages, we still observe a positive relationship between other measures of FACB
and wage compliance. In addition, our index of Internal Relations and Benefits carries the
largest positive coefficient, also significant at the 1% level, despite the fact that this category
contains some of the measures most likely to be traded off with wages (benefits).
In contrast, the OSH (Occupational Safety and Health) coefficient is positive but
insignificant, suggesting that, if firms are trading off any form of working conditions with wage
24
compliance, this category may represent them. Nonetheless, because this category’s
insignificance differs so strongly from the results found earlier in this section, we explore OSH in
greater depth. Column two of Table 5c shows regression results for Equation (3), with the
smaller subcategory components of OSH substituted in for the broader category variable. The
results, a list of insignificant coefficients hovering around zero, fail to reveal any hidden
relationships within OSH, instead confirming the lack of a significant relationship between wage
compliance and OSH.
While the disaggregation of OSH failed to turn up any hidden relationships, it might be
that the disaggregation itself was the problem. Empirically, multicollinearity could be the issue,
and theoretically, such relationships may only emerge with more aggregate variables because of
a firm’s holistic approach to choosing a package of working conditions to offer. For this reason,
and to provide more a more detailed analysis of the other categories, we disaggregate FACB and
Internal Relations and Benefits. When we split FACB, we find that two of the three
subcategories (Unions and Strikes) carry statistically significant positive coefficients, while the
third (Shop Stewards) is insignificant. These results give no indication of a multicollinearity
issue caused by disaggregation.
To divide Internal Relations and Benefits, we first split it into Benefits and Internal
Relations, with the results shown in column four of Table 5b. Even this relatively small change
in specification is revealing, as the Benefits coefficient is statistically insignificant, consistent
with the expectation that firms would be more likely to trade off benefits and wages. The
Internal Relations coefficient remains relatively large and statistically significant. To provide an
even more detailed picture and to further test the multicollinearity question, we further
disaggregate both Benefits and Internal Relations in columns five and six (respectively) of Table
25
5b. Disaggregation of Benefits yields no coefficients that statistically differ from zero, consistent
with the Benefits coefficient as a whole. Disaggregation of Internal Relations, meanwhile,
reveals that Core Standards and Working Time are statistically significantly related to wages.
Furthermore, it appears that Core Standards is largely responsible for the magnitude of the
Internal Relations aggregated coefficient, though Working Time appears to play an important
role in its significance. The statistical significance of these results does indicate that
multicollinearity plays at most a minimal role, suggesting that the insignificance of OSH in
predicting wage compliance may reflect a true zero relationship between the two. A zero
relationship is still non-negative, though, so we continue to fail to find evidence supporting
compensating differential theory within these foreign-exposed firms.
6. Foreign Ownership and the Wage-Working Conditions Relationship
6.1 Initial Results
To determine how wage compliance and working conditions are differently related in
foreign-owned firms than in domestically owned ones, we add a foreign ownership dummy
variable and that dummy interacted with Working Conditions (WC) to Equation (2a) to get
Equation (4) below:
∆Wages i(t-[t-1])= ß0 + ß1(∆WC i(t-[t-1]))+ ß2(∆FirmSize i(t-[t-1]))+ ß3(FirmAgeit)+ ß4(∆%Union i(t-[t-1]))
+ ß5(Visitit)+ ß6(∆Time i(t-[t-1]))+ ß7(Foreign-Ownedit)+ ß8(Foreign-Ownedit*∆WC i(t-[t-1]))+ εit (4)
With this specification, the coefficient on the Working Conditions variable represents the
relationship between wage compliance and working conditions in domestically owned firms,
while the interaction term’s coefficient represents the marginal impact of foreign ownership on
that relationship. Adding ß1 and ß8, therefore, gives the total impact of working conditions on
wage compliance in foreign-owned firms. Initial results for Equation (4), shown in the first
26
column of Table 6a, look very similar to those in Table 5a. R-squared remains low at 0.10, and
most controls’ coefficients remain statistically insignificant and small. Interestingly, the
Working Conditions variable maintains a positive and statistically significant coefficient, and its
magnitude nearly triples, indicating that the positive relationship between wage compliance and
working conditions is stronger in the domestically owned firms than in the sample as a whole. In
these domestically owned firms, when Working Conditions improve by ten percentage points,
wage compliance improves by about 24 percentage points, a very large effect.
The negative coefficient on the Foreign Ownership x Working Conditions interaction
term, meanwhile, suggests that marginal impact of foreign ownership on the wage-working
conditions relationship is negative. The total effect of working conditions on wage compliance
in foreign-owned firms is positive and statistically significant at the 1% level, but the effect is
much smaller (an 8-percentage-point increase in wage compliance for a 10-percentage-point
improvement in working conditions) than that in domestically owned firms. Given that
compliance on both wages and working conditions is higher in foreign-owned firms, the smaller
positive relationship in these firms is unsurprising; beyond some high level of compliance,
additional improvements in wage and/or working conditions compliance become less feasible
and the marginal effort returns on these improvements may diminish.
6.2 Robustness
Columns two through six of Table 6a show results for a variety of different specifications
and sample changes, most of which are identical to those reported in section five. As before, the
Working Conditions coefficient changes little with the varying specifications, and the Foreign
Ownership and interaction coefficients generally remain fairly stable as well. Columns three and
four of Table 6a show results with unionization excluded and a time variable added, respectively.
27
The pattern of positive wage-working conditions relationships in all firms (but a stronger effect
of working conditions on wage compliance in domestically owned firms) remains through these
specification changes.
The positive relationship also remains when we control for the monitors that visited the
factory (column two of Table 6a), but the marginal negative effect of foreign ownership becomes
statistically insignificant in this specification. These results correspond interestingly with the
results shown in columns five and six of Table 6a, in which we split the sample into the original
and IMS firms. In the IMS firms, the statistical significance of the foreign ownership impact on
the wage-working conditions relationship disappears, but the impact of foreign ownership is
much stronger in the original firms. Because there was incomplete overlap in monitors between
the two time periods, some monitors are present only for the first set of visits to the original
firms, so the monitor controls in the results presented in column two of Table 6a could be
capturing the same effect as the contrast between columns five and six – a distinct marginal
effect of foreign ownership between these two samples. These results continue to confirm the
positive wage-working conditions relationship in both domestically and foreign-owned firms, but
present a potential caveat to the conclusion that foreign ownership reduces the strength of the
wage-working conditions relationship in these firms.
6.3 Disaggregated Working Conditions and Foreign Ownership
The results presented in Table 6a focus on working conditions and Foreign Ownership
variables that are both aggregated for simplicity, but given the detailed data we have available,
we can also disaggregate these variables into their components. First, as shown in section 5.4,
we can disaggregate the Working Conditions variable into four groups of working conditions.
Replacing the Working Conditions variable with these four smaller variables and interacting each
28
of these smaller variables with Foreign Ownership yields the results shown in Table 6b. The
results serve to clarify somewhat the difference between the wage-working conditions
relationship in domestically owned firms (the stand-alone working conditions coefficients in the
first column) and the relationship in foreign-owned firms (the total effect coefficients in the third
column). In domestically owned firms, Paperwork and Internal Relations and Benefits are
significantly positively related to wage compliance, while we find some evidence of
compensating differentials in the statistically significant negative coefficient on FACB (Freedom
of Association and Collective Bargaining). In foreign-owned firms, we find no evidence of
compensating differentials, but we find weak positive relationships of wage compliance with
Paperwork and FACB. Consistent with the results with the aggregated Working Conditions
variable, we generally find foreign ownership to weaken but not eliminate the positive effect
between wage compliance and working conditions.
The differing effect of FACB in the two groups is an interesting exception to this general
finding, especially because it is the only working conditions measure for which we find
statistically significant evidence of a compensating differential relationship. Surprisingly, given
the consistently weaker positive wage-working conditions relationship in foreign-owned firms,
we find this isolated evidence of compensating differentials in domestically owned firms. In this
case, foreign ownership has a large positive impact on the wage-working conditions relationship,
an impact large enough to produce a total working conditions effect that is statistically
significant and positive. This interesting result certainly warrants further exploration of the
relationship between Freedom of Association/Collective Bargaining and wages, but we will
leave this task for future research.
29
The impact of foreign ownership on wage compliance might vary by the source country
in addition to varying by the category of working conditions considered. The results in Table 6c
explore this possibility by including a set of country of ownership dummies (using the countries
and groups shown in Table 3c) and their interactions with Working Conditions. As before,
working conditions (measured again as the aggregate Working Conditions variable) are
significantly positively related to wage compliance in domestically owned firms. The interaction
terms are all negative and most are statistically significant (with the exceptions of China and
Other Asia), affirming the general result that foreign-owned firms exhibit a smaller positive
wage-working conditions relationship than domestically owned firms. Furthermore, the
disaggregated ownership variables reveal that, in some cases, the wage-working conditions
relationship is statistically indistinct from zero. In no case, however, do we observe a
statistically significant negative relationship between wage compliance and working conditions.
The broad overview of these results thus provides further evidence of a non-negative relationship
between wages and working conditions, while confirming this relationship’s statistically
significant variation between domestically and foreign-owned firms.
While the results generally support the findings of Section 6.1, the variation in the
interaction term coefficients illustrates that the effect of foreign ownership on the wage-working
conditions relationship differs by the source country. Firms from the West, Korea, Malaysia, and
Singapore all have a statistically significantly (5% level) smaller positive relationship between
wage compliance and working conditions, relative to Cambodian firms. In contrast to the
aggregated foreign ownership results, the interaction effects yield a total wage-working
conditions relationship that is not statistically significantly positive in these firms. Though the
disaggregated interaction terms do not reveal any powerful hidden evidence of compensating
30
differentials, these results do show that firms in these countries exhibit no relationship at all
between wage compliance and working conditions.
Meanwhile, firms from China, Hong Kong, Taiwan, and the other Asian country group
held a positive and statistically significant relationship between wage compliance and working
conditions, consistent with the results found with the aggregated foreign ownership variable.18
These results indicate a greater similarity in patterns of compliance between Cambodian firms
and those affiliated with China (firms from China, Hong Kong, and Taiwan) than between
Cambodian firms and the rest of the firms. These varying relationships are fascinating and
should be the topic of more in-depth future research. Our fundamental point, however, remains
that only for one country and one measure of working conditions measure do we see any
evidence of compensating differentials. In the vast majority of scenarios, working conditions
and wage compliance are positive related in all firms, but more so in domestically owned firms.
7. Conclusion
We have shown, first of all, that compliance on both wages and working conditions is
higher in foreign-owned firms, contradicting the compensating differentials explanation for
foreign ownership wage premiums. Furthermore, in this sample of Cambodian exporting
garment factories as a whole, wage compliance and working conditions are positively related,
supporting an efficiency wages explanation of why some firms pay higher wages than others and
indicating that workers are made better off overall by working in firms that pay them higher
wages. This positive wage-working conditions relationship, while smaller in foreign-owned
firms as a whole, also suggests that both domestically and foreign-owned firms in this sample
18
This positive overall relationship emerges in Hong Kong and Taiwan despite a statistically significantly (10%
level) smaller positive relationship in these countries’ firms relative to Cambodian firms. In other words, while they
maintain a positive and statistically significant overall relationship between wage compliance and working
conditions, the relationship is statistically significantly smaller in these firms than in Cambodian firms.
31
have responded to a positive working conditions shock by increasing the worker compensation
package overall, thereby shifting their effort curves out. This finding implies that programs like
Better Factories Cambodia can push for improvements in working conditions without inducing a
reduction in wage compliance, so such programs might increase overall worker welfare.
We present these results with reservation, however, due to some fundamental weaknesses
in our dataset and results. First and most importantly, the sample size of domestically owned
firms is quite small relative to foreign-owned firms. Due to this small sample size, our results
may not be generally applicable for Cambodian firms, let alone firms in any other country. In
addition, our sample contains no firms that change ownership from domestic to foreign or vice
versa during the sampling period. As a result, we must rely on a between-firms assessment of
the foreign ownership effect, preventing us from taking a true ceteris paribus look at the foreign
ownership effect on the wage-working conditions relationship. Finally, our empirical results are
characterized by low r-squared values that indicate a failure to effectively predict wage
compliance using our control variables. Undoubtedly, the ideal regression would contain
additional control variables to improve the explanatory power of the independent variables, but
we face a less-than-ideal (though uniquely comprehensive) dataset. In essence, we analyze a
limited sample of domestic firms, with no within-firm variation in ownership, and explain only
about 10% of the variation in wage compliance using our explanatory variables. With that said,
we also acknowledge that our results are robust to a range of specification alterations aimed at
correcting or at least exposing these weaknesses.
This body of research, furthermore, is by no means complete. We present only a single-
sector, single-country, single-dimension case study of globalization’s effect on the wage-working
conditions relationship. As the ILO’s Better Work program extends the Better Factories
32
Cambodia model to other developing countries, further research can address this question on a
multi-country scale across sectors and including non-exporting firms for broader applicability of
results. The BFC dataset itself also contains the potential for further research to expand our
understanding of the wage-working conditions relationship. First of all, the interesting findings
above of differing wage-working conditions relationships between working conditions measures
and source countries provides an excellent opportunity for additional understanding of this
complex issue. Meanwhile, while our categorizations of working conditions make sense in the
way they affect workers, they may not accurately reflect the cost analysis in the firm (for
example, Occupational Safety and Health measures are grouped together but the costs of
improving these measures can vary widely). Alternate categorizations of the working conditions
measures might therefore give a clearer picture of the wage-working conditions relationship and
how it varies among different measures. Finally, assessing the pair-wise relationships between a
variety of different working conditions measures with one another could also reveal more about
how firms make decisions in their provision of working conditions for workers. While this sort
of analysis is beyond the scope of this paper, it is well within the means of this rich dataset.
33
References
Aitken, Brian, Ann Harrison, and Robert E. Lipsey. 1996. Wages and Foreign Ownership: A
Comparative Study of Mexico, Venezuela, and the United States. Journal of International
Economics 40, no. 3-4:345-371.
Altonji, Joseph G., and Christina H. Paxson. 1988. Labor Supply Preferences, Hours Constraints,
and Hours-Wage Trade-Offs. Journal of Labor Economics 6, no. 2:254-276.
Arai, Mahmood. 1994. An Empirical Analysis of Wage Dispersion and Efficiency Wages.
Scandinavian Journal of Economics 96, no. 1:31-50.
Bernard, Andrew B., and J. B. Jensen. 1995. Exporters, Jobs, and Wages in US Manufacturing -
1976-1987. Brookings Papers on Economic Activity.
Brown, Charles, and James L. Medoff. 2003. Firm age and wages. Journal of Labor Economics
21, no. 3:677-697.
Brown, Charles. 1980. Equalizing Differences in the Labor-Market. Quarterly Journal of
Economics 94, no. 1:113-134.
Brown, Charles, and James Medoff. 1989. The Employer Size-Wage Effect. Journal of Political
Economy 97, no. 5:1027-1059.
Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern. 2002. The Effects of Multinational
Production on Wages and Working Conditions in Developing Countries. University of
Michigan Discussion Paper No. 483.
Cappelli, Peter, and Keith Chauvin. 1991. An Interplant Test of the Efficiency Wage Hypothesis.
Quarterly Journal of Economics 106, no. 3:769-787.
Cousineau, Jean-Michel, Robert Lacroix, and Anne-Marie Girard. 1992. Occupational Hazard
and Wage Compensating Differentials. Review of Economics and Statistics 74, no. 1:166-
169.
Dorman, Peter, and Paul A. Hagstrom. 1998. Wage Compensation for Dangerous Work
Revisited. Industrial and Labor Relations Review 52, no. 1:116-135.
Duncan, Greg J., and Bertil Holmlund. 1983. Was Adam Smith Right after All? Another Test of
the Theory of Compensating Wage Differentials. Journal of Labor Economics 1, no. 4:366-
379.
Duncan, Greg J., and Stafford, Frank P. 2002 [1980]. Do Union Members Receive
Compensating Wage Differentials? In The Economics of Labor Unions. Volume 1, edited by
Alison L. Booth. Cheltenham, U.K. and Northampton, Mass.: Elgar Reference Collection.
International Library of Critical Writings in Economics, vol. 147.
Fuess, Scott M., and Meghan Millea. 2002. Do Employers Pay Efficiency Wages? Evidence
from Japan. Journal of Labor Research 23, no. 2:279-292.
Girma, Sourafel, and Holger Gorg. 2007. Evaluating the Foreign Ownership Wage Premium
Using a Difference-in-Differences Matching Approach. Journal of International Economics
72, no. 1:97-112.
Glick, Peter, and Francois Roubaud. 2006. Export Processing Zone Expansion in Madagascar:
What Are the Labour Market and Gender Impacts? Journal of African Economies 15, no.
4:722-756.
Goldsmith, Arthur H., Jonathan R. Veum, and William A. Darity. 2000. Working Hard for the
Money? Efficiency Wages and Worker Effort. Journal of Economic Psychology 21, no.
4:351-385.
Konings, Jozef, and Patrick P. Walsh. 1994. Evidence of Efficiency Wage Payments in UK Firm
Level Panel-Data. Economic Journal 104, no. 424:542-555.
Leigh, J. P. 1991. No Evidence of Compensating Wages for Occupational Fatalities. Industrial
Relations 30, no. 3:382-395.
34
Lipsey, R. E. 2004. Home- and host-country effects of foreign direct investment. Edited by RE
Baldwin, L. A. Winters. Stockholm, SWEDEN ed.
Mandel, Benjamin R. 2008. Heterogeneous Firms and Import Quality: Evidence from
Transaction-Level Prices. September 2008: Economics Dept., University of California
Davis.
Marin, Alan, and George Psacharopoulos. 1982. The Reward for Risk in the Labor-Market -
Evidence from the United-Kingdom and a Reconciliation with Other Studies. Journal of
Political Economy 90, no. 4:827-853.
McCrate, Elaine. 2005. Flexible Hours, Workplace Authority, and Compensating Wage
Differentials in the US. Feminist Economics 11, no. 1:11-39.
McNabb, Robert. 1989. Compensating Wage Differentials: Some Evidence for Britain. Oxford
Economic Papers, N. S. 41, no. 2:327-338.
Robertson, Raymond E., and Neak, Samsen. 2008. The Effects of Globalization on Working
Conditions: Cambodia. Working Paper ed. World Bank.
Robertson, Raymond E., et al. 2008. Globalization and Working Conditions: Evidence from
Indonesia. Working Paper ed. World Bank.
Robertson, Raymond E., and Trigueros-Argüello, Alvaro. 2008. The Effects of Globalization on
Working Conditions: El Salvador 1995-2005.Working Paper ed.World Bank.
Schank, Thorsten, Claus Schnabel, and Joachim Wagner. 2007. Do Exporters Really Pay Higher
Wages? First Evidence from German Linked Employer-Employee Data. Journal of
International Economics 72, no. 1:52-74.
Sjoholm, Fredrik, and Robert E. Lipsey. 2006. Foreign Firms and Indonesian Manufacturing
Wages: An Analysis with Panel Data. Economic Development and Cultural Change 55, no.
1:201-221.
35
Table 3a: Contents of Aggregated Working Conditions
Variables
Wages (5)
Minimum Wage; Premium Wages for Night Work,
Overtime, Holiday Work, and Work on Weekly Time Off
Working
Conditions (127)
OSH, Internal Relations and Benefits, Paperwork, FACB
(see below)
OSH (43)
Occupational Health and Safety: Health Facilities; Water and
Toilet; Temperature, Ventilation, Noise, and Lighting;
Machine Safety; Safety of Operations and Workplace
Motion; Emergency Preparedness; Chemical Safety
Internal
Relations
(23)
Child Labor, Discrimination, Forced Labor,
Discipline/Management Conduct, Overtime, Regular Hours,
Weekly Rest, Liaison Officers, Internal Disputes
Internal
Relations and
Benefits (38)
Benefits
(15)
Holiday, Annual, and Special Leave; Worker's
Compensation; Maternity Leave and Benefits
Paperwork (33)
Informing Workers about Wages/Holidays/Working Time,
Internal Regulations, Contracts/Hiring Procedures,
Collective Agreements, MOSALVY (Cambodian Labor
Ministry) Reporting/Permissions, Chemical Documentation,
Health and Safety Assessment and Reporting
FACB (13)
Freedom of Association and Collective Bargaining: Unions,
Strikes, Shop Stewards Notes: Number of questions contained in the index shown in parentheses. Listed contents of Wages variable are all
individual questions, while listed contents of all other variables are groups of questions.
36
Table 3b: Summary Statistics
Variable Obs
Mean/
%
Std.
Dev. Min Max
Firm Age (Years) 614 4.79 2.56 0.58 14.08
D. Firm Age 614 0.84 0.86 0.08 5.08
Firm Size (100s of Workers) 614 12.06 11.13 0.16 75.12
∆ Firm Size 614 0.41 3.00 -13.51 30.52
% Union (% Workers) 614 40.22 32.26 0.00 136.16
∆ % Union 614 4.93 24.56 -102.55 102.32
Visit (#) 614 3.07 0.96 2.00 6.00
Time Difference (Days) 614 10.26 10.48 0.70 62.57
Wage Compliance (%) 614 91.82 18.25 0.00 100.00
∆ Wage Compliance 614 2.28 16.59 -80.00 80.00
Working Conditions Compliance (%) 614 85.69 6.50 62.99 97.64
∆ Working Conditions 614 1.76 4.91 -14.17 35.43
Paperwork Compliance (%) 614 87.05 8.81 54.55 100.00
∆ Paperwork Compliance 614 2.36 6.63 -24.24 30.30
FACB Compliance (%) 614 90.54 7.25 53.85 100.00
∆ FACB Compliance 614 1.23 8.71 -23.08 46.15
IR/Benefits Compliance 614 87.66 6.50 63.16 100.00
∆ IR/Benefits Compliance 614 1.64 6.11 -18.42 23.68
OSH Compliance (%) 614 81.44 9.96 37.21 100.00
∆ OSH Compliance 614 1.57 7.53 -25.58 62.79
37
Table 3c: Countries of Ownership
Country
Entire
Sample
% of Entire
Sample
Firms with
2+ Visits
% of Firms
with 2+ Visits
Taiwan 87 24.6% 76 26.3%
Hong Kong SAR 76 21.5% 57 19.7%
China 70 19.8% 55 19.0%
China 69 54
Macau SAR 1 1
Korea 40 11.3% 33 11.4%
Malaysia 19 5.4% 19 6.6%
Singapore 15 4.2% 13 4.5%
West 14 4.0% 10 3.5%
American Samoa 1 1
Australia 4 2
Canada 1 1
France 1 0
Germany 1 0
United Kingdom 2 2
United States 4 4
Other Asia 6 1.7% 6 2.1%
Bangladesh 1 1
Indonesia 2 2
Philippines 1 1
Thailand 1 1
Viet Nam 1 1
Cambodia 27 7.6% 20 6.9%
38
Table 3d: Wage and Working Conditions Compliance by FDI
Variable Obs
Mean
(All
Visits)
Std.
Dev. Min Max
Mean
(Visit
1)
Mean
(Visits
4-5)
Wage Compliance (%) 614 91.82 18.25 0.00 100.00 88.49 95.12
Wage Compliance in Foreign-Owned (%) 582 90.00 19.13 0.00 100.00 84.75 95.13
Wage Compliance in West-Owned (%) 17 91.76 14.25 60.00 100.00 87.50 100.00
Wage Compliance in China-Owned (%) 114 84.04 24.41 0.00 100.00 74.63 93.75
Wage Compliance in Hong Kong-Owned (%) 113 91.86 18.05 20.00 100.00 85.14 96.82
Wage Compliance in Singapore-Owned (%) 27 93.33 17.54 20.00 100.00 88.89 97.78
Wage Compliance in Taiwan-Owned (%) 182 90.33 18.17 0.00 100.00 88.85 93.13
Wage Compliance in Korea-Owned (%) 70 90.57 18.25 20.00 100.00 85.38 95.45
Wage Compliance in Malaysia-Owned (%) 45 95.56 10.35 60.00 100.00 92.31 97.78
Wage Compliance in Other Asia-Owned (%) 14 90.00 17.10 40.00 100.00 80.00 96.00
Wage Compliance in Domestically Owned (%) 32 81.25 30.87 0.00 100.00 78.57 85.00
∆ Wage Compliance 614 2.28 16.59 -80.00 80.00 4.15 0.39
∆Wage Compliance in Foreign-Owned (%) 582 2.44 15.92 -80.00 80.00 4.75 0.41
∆Wage Compliance in West-Owned (%) 17 2.35 6.64 0.00 20.00 2.50 0.00
∆Wage Compliance in China-Owned (%) 114 4.91 20.71 -80.00 80.00 9.76 -2.50
∆Wage Compliance in Hong Kong-Owned (%) 113 1.59 16.51 -60.00 60.00 4.57 -0.45
∆Wage Compliance in Singapore-Owned (%) 27 0.74 8.74 -20.00 40.00 4.44 0.00
∆Wage Compliance in Taiwan-Owned (%) 182 0.88 14.54 -60.00 60.00 0.66 0.94
∆Wage Compliance in Korea-Owned (%) 70 4.86 16.83 -20.00 80.00 6.92 3.64
∆Wage Compliance in Malaysia-Owned (%) 45 1.33 8.94 -20.00 20.00 3.08 1.11
∆Wage Compliance in Other Asia-Owned (%) 14 4.29 13.99 -20.00 40.00 12.00 4.00
∆Wage Compliance in Domestically Owned
(%) 32 -0.63 26.14 -80.00 60.00 -4.29 0.00
Working Conditions (WC) Compliance (%) 614 85.69 6.5 62.99 97.64 84.22 87.19
WC Compliance in Foreign-Owned (%) 582 85.94 6.32 62.99 97.64 84.5 87.34
WC Compliance in West-Owned (%) 17 86.48 5.75 77.95 96.85 85.33 88.19
WC Compliance in China-Owned (%) 114 80.74 6.52 58.27 93.7 78.03 83.54
WC Compliance in Hong Kong-Owned (%) 113 84.02 8.09 60.63 97.64 80.11 87.24
WC Compliance in Singapore-Owned (%) 27 85.39 7.59 67.72 96.06 81.19 88.98
WC Compliance in Taiwan-Owned (%) 182 85.12 6.64 66.93 96.85 82.7 87.4
WC Compliance in Korea-Owned (%) 70 85.04 6.11 72.44 95.28 83.53 86.69
WC Compliance in Malaysia-Owned (%) 45 88.17 4.91 75.59 96.85 85.22 90.64
WC Compliance in Other Asia-Owned (%) 14 81.5 6.3 68.5 89.76 77.95 85.67
WC Compliance in Domestically Owned (%) 32 79.4 8.12 66.93 93.7 78.12 82.87
∆ Working Conditions 614 1.76 4.91 -14.17 35.43 3.14 0.35
∆WC in Foreign-Owned (%) 582 1.76 4.94 -14.17 35.43 3.21 0.34
∆WC in West-Owned (%) 17 2.04 4.32 -7.87 11.02 4.43 -0.26
∆WC in China-Owned (%) 114 2.16 5.55 -7.87 35.43 3.28 0.94
∆WC in Hong Kong-Owned (%) 113 1.79 5.16 -11.81 19.69 3.22 0.39
∆WC in Singapore-Owned (%) 27 1.60 5.11 -6.30 15.75 6.12 -0.70
∆WC in Taiwan-Owned (%) 182 1.33 4.89 -14.17 18.90 2.65 -0.11
∆WC in Korea-Owned (%) 70 1.69 3.90 -7.09 12.60 3.06 0.72
∆WC in Malaysia-Owned (%) 45 1.96 4.52 -7.87 15.75 2.67 0.70
∆WC in Other Asia-Owned (%) 14 3.43 5.21 -7.09 11.81 4.57 1.10
∆WC in Domestically Owned (%) 32 1.82 4.53 -5.51 11.81 2.08 0.69
39
Table 4a: Regression Results – Foreign Ownership and
Wages/Working Conditions 1 2 3 4
Wages (1) Wages (2)
Working
Conditions (1)
Working
Conditions (2)
Foreign 9.220 9.392 4.317 2.667
Ownership (5.599)* (3.955)** (1.518)*** (1.343)**
Firm Age 1.143 -0.052 0.977 -0.500
(Years) (0.321)*** 0.424 (0.107)*** (0.144)***
Firm Size 0.208 0.244 0.132 0.214
(100s of Workers) (0.090)** (0.096)** (0.038)*** (0.032)***
Unionization 0.032 0.046 0.008 0.018
(% Workers) (0.024) (0.037) (0.008) (0.013)
Constant 72.590 76.493 74.211 80.673
(5.499)*** (4.331)*** (1.548)*** (1.470)***
Observations 936 936 936 936
Firms 288 288 288 288
R2 0.061 0.06 0.311 0.18
• significant at 10%; ** significant at 5%; *** significant at 1% 1 R-squared within
Robust standard errors in parentheses for columns one and three; columns two and four use an empirical method that
does not permit robust standard error calculation. Regression results: Eq. 1, wages as the dependent variable with
random effects (column 1) and between effects (column 2); and working conditions as the dependent variable with
random effects (column 3) and between effects (column 4).
40
Table 4b: Regression Results – Disaggregated Foreign Ownership
and Wages/Working Conditions
1 2 3 4
Wages (1) Wages (2)
Working
Conditions (1)
Working
Conditions (2)
Firm Age 1.174 -0.049 0.987 -0.48
(Years) (0.328)*** (0.438) (0.108)*** (0.145)***
Firm Size 0.17 0.194 0.108 0.186
(100s of Workers) (0.095)* (0.100)* (0.039)*** (0.033)***
Unionization 0.034 0.049 0.008 0.016
(% Workers) (0.024) (0.038) (0.008) (0.012)
West 8.965 8.157 4.58 2.25
(6.592) (6.489) (2.139)** (2.144)
China 5.369 5.135 1.342 0.017
(6.089) (4.417) (1.701) (1.459)
Hong Kong 9.353 10.794 3.506 3.111
(6.016) (4.403)** (1.698)** (1.455)**
Taiwan 9.152 9.393 5.368 3.615
(5.821) (4.307)** (1.592)*** (1.423)**
Korea 13.149 12.261 6.363 3.803
(6.099)** (4.804)** (1.724)*** (1.587)**
Malaysia 14.456 14.473 8.043 5.887
(5.956)** (5.490)*** (1.809)*** (1.814)***
Singapore 11.425 11.416 4.72 2.276
(6.485)* (6.122)* (2.222)** (2.023)
Other Asia 8.978 10.871 1.951 2.569
(7.537) (7.749) (2.806) (2.560)
Constant 72.668 76.712 74.426 80.798
(5.527)*** (4.369)*** (1.536)*** (1.443)***
Observations 936 936 936 936
Firms 288 288 288 288
R2 0.061 0.08 0.311 0.24
• significant at 10%; ** significant at 5%; *** significant at 1% 1 R-squared within
Robust standard errors in parentheses for columns one and three; columns two and four use an empirical method that
does not permit robust standard error calculation. Regression results: Eq. 1, wages as the dependent variable with
random effects (column 1) and between effects (column 2); and working conditions as the dependent variable with
random effects (column 3) and between effects (column 4).
41
Table 5a: Regression Results – Aggregated Working
Conditions
1 2 3 4 5 6
Firm Age -3.512 -0.464 0.161 0.402 -0.068 0.196
(Years) (5.116) (1.160) (0.276) (0.280) (0.316) (0.289)
Firm Size 0.251 0.188 0.427 0.506 0.342 0.431
(100s of Workers) (0.286) (0.242) (0.257)* (0.280)* (0.238) (0.258)*
Unionization 0.019 0.019 0.04 0.042 0.041
(% Workers) (0.036) (0.034) (0.034) (0.033) (0.034)
Visit # 3.067 1.273 -0.552 -1.435 -0.244 -0.777
(3.099) (0.878) (0.738) (0.798)* (0.735) (0.877)
Time Between 0.035 -0.098 -0.232 0.014 -0.034
Visits (Months) (0.085) (0.123) (0.143) (0.081) (0.195)
Working 0.783 1
0.873 1
0.869 0.891 0.802 0.875
Conditions (0.243)*** (0.194)*** (0.204)*** (0.217)*** (0.172)*** (0.209)***
Time 0.61
(Years) (1.582)
Constant 27.974 12.536 1.888 21.684 1.801 -1221.638
(27.764) (14.600) (2.080) (22.606) (1.977) (3174.486)
Observations 614 981 614 614 769 614
Firms 278 333 278 278 289 278
R-Squared 0.08 0.13 0.08 0.15 0.06 0.08 * significant at 10%; ** significant at 5%; *** significant at 1%
1 = Working Conditions variable in levels (not differences)
Robust standard errors in parentheses. Regression results for Eq.2 (column 1), Eq. 2 with Time Between Visits
excluded (2), Eq. 2a (3), Eq. 2a with monitor controls (4), Eq. 2a with unionization excluded (5), and Eq. 2a with a
continuous time control (6). Reported R2 values are R
2 within.
42
Table 5b: Regression Results – Aggregated Working
Conditions (Continued) 1 2 3 4 5 6
Firm Age (Years) -0.474 0.568 0.154 0.065 0.214 -0.049
(0.799) (0.336)* (0.240) (0.555) (0.409) (0.544)
Firm Size 1.079 0.097 0.226 0.808 0.146 0.771
(100s of Workers) (0.670) (0.310) (0.291) (0.487)* (0.212) (0.496)
Unionization 0.037 0.038 -0.007 0.089 0.007 0.065
(% Workers) (0.082) (0.037) (0.045) (0.049)* (0.025) (0.053)
Visit # 1.982 -2.043 -0.553 -0.652 -0.245 -0.279
(1.831) (0.890)** (0.729) (1.496) (0.861) (1.518)
Time Between -0.074 -0.944 0.101 -0.205 0.035 -0.149
Visits (Months) (0.190) (0.337)*** (0.098) (0.159) (0.095) (0.157)
Working 0.762 0.892 0.436 1.141 0.924 0.844
Conditions (0.361)** (0.240)*** (0.185)** (0.303)*** (0.286)*** (0.275)***
Constant -1.154 9.902 0.604 3.484 -0.385 3.144
(8.789) (3.517)*** (2.423) (3.224) (2.745) (3.331)
Observations 163 451 306 308 313 301
Firms 71 207 130 148 162 165
R-Squared 0.06 0.10 0.06 0.11 0.21 0.08
* significant at 10%; ** significant at 5%; *** significant at 1%
Regression results for Eq.2a for original factories only (column 1); Eq. 2a for IMS factories (2); Eq. 2a for high-
compliance firms, >85% (3); Eq. 2a for low-compliance firms, <85% (4); Eq. 2a for high-compliance observations,
>87% (5); and Eq. 2a for low-compliance observations, <87% (6). Robust standard errors in parentheses. Reported
R2 values are R
2 within.
43
Table 5c: Regression Results – Disaggregated Working
Conditions Variables
1 2 3 4 5 6
Firm Age (Years) 0.213 0.211 0.212 0.216 0.221 0.192
(0.272) (0.267) (0.271) (0.274) (0.274) (0.269)
Firm Size 0.417 0.430 0.416 0.422 0.421 0.426
(100s of Workers) (0.253)* (0.250)* (0.249)* (0.254)* (0.253)* (0.255)*
Unionization 0.041 0.041 0.04 0.042 0.042 0.036
(% Workers) (0.033) (0.033) (0.033) (0.033) (0.033) (0.033)
Visit # -0.572 -0.494 -0.626 -0.573 -0.559 -0.537
(0.737) (0.711) (0.743) (0.737) (0.742) (0.740)
Time Between -0.122 -0.153 -0.15 -0.117 -0.136 -0.109
Visits (Months) (0.126) (0.137) (0.128) (0.129) (0.135) (0.131)
Paperwork 0.330 0.359 0.293 0.329 0.324 0.329
(0.126)*** (0.130)*** (0.123)** (0.126)*** (0.126)** (0.124)***
OSH 0.105 See 0.134 0.105 0.102 0.1
(0.152) Table 5d1 (0.151) (0.152) (0.152) (0.152)
FACB 0.181 0.188 See 0.181 0.186 0.168
(0.096)* (0.095)** Table 5d1 (0.096)* (0.099)* (0.096)*
Internal Relations 0.362 0.355 0.349
and Benefits (0.136)*** (0.142)** (0.133)***
Benefits 0.121 See 0.127
(0.090) Table 5d1 (0.092)
Internal Relations 0.239 0.24 See
(0.100)** (0.100)** Table 5d1
Constant 1.740 1.672 2.084 1.688 1.769 1.880
(2.078) (2.061) (2.083) (2.105) (2.110) (2.112)
Observations 614 614 614 614 614 614
R-squared Within 0.09 0.09 0.09 0.09 0.09 0.10 * significant at 10%; ** significant at 5%; *** significant at 1%
Robust standard errors in parentheses. Regression results for Eq. 3 (column 1), Eq. 3 with OSH split (2), Eq. 3 with
FACB split (3), Eq. 3 with Working Time/Core/Benefits split into Working Time/Core and Benefits (4), Eq. 3 with
Benefits split (5), and Eq. 3 with Working Time/Core Standards split (6). Reported R2 values are R
2 within.
Coefficients of divided categories are shown in Table 5d below.
44
Table 5d: Regression Results – Disaggregated Working
Conditions Variables (Continued, Subcategory Coefficients)
Category Subcategory 2 3 5 6
Health/First Aid 0.000
(0.057)
Machine Safety 0.047
(0.120)
Temp/Vent/ 0.032
Noise/Light (0.052)
Welfare Facilities 0.001
(0.058)
Operations/ 0.094
Physical Plant (0.083)
Emergency -0.020
Preparedness (0.056)
Chemical Safety -0.029
OSH
(0.026)
Strikes 0.231
(0.119)*
Unions 0.21
(0.122)*
Shop Stewards 0.010
FACB
(0.033)
Workers' 0.089
Compensation (0.073)
Leave/Holidays 0.018
(0.054)
Maternity Benefits 0.033
Benefits
(0.056)
Disputes -0.029
(0.048)
Management 0.011
Conduct (0.036)
Working Time 0.093
(0.047)**
Liaison Officer -0.012
(0.050)
Core Standards 0.274
Core/
Working
Time
(0.152)* * significant at 10%; ** significant at 5%; *** significant at 1%
Robust standard errors in parentheses. Regression results for Eq. 3 with OSH split (2), Eq. 3 with FACB split (3),
Eq. 3 with Benefits split (5), and Eq. 3 with Working Time/Core Standards split (6). Reported R2 values are R
2
within.
45
Table 6a: Regression Results – Aggregated Foreign Ownership
and Working Conditions 1 2 3 4 5 6
Firm Age (Years) 0.167 0.381 -0.033 0.204 -0.210 0.568
(0.272) (0.281) (0.315) (0.280) (0.714) (0.338)*
Firm Size 0.420 0.506 0.327 0.424 0.954 0.101
(100s of Workers) (0.256) (0.275)* (0.239) (0.256)* (0.646) (0.311)
Unionization 0.038 0.041 0.039 0.060 0.037
(% Workers) (0.033) (0.033) (0.034) (0.074) (0.036)
Visit # -0.629 -1.424 -0.339 -0.865 1.217 -2.043
(0.735) (0.801)* (0.733) (0.867) (1.819) (0.892)**
Time Between -0.088 -0.218 0.023 -0.021 -0.040 -0.966
Visits (Months) (0.119) (0.141) (0.080) (0.189) (0.185) (0.348)***
Working 2.319 2.190 2.107 2.321 5.014 1.203
Conditions (0.855)*** (0.954)** (0.711)*** (0.852)*** (1.892)*** (0.580)**
Time 0.638
(Years) (1.556)
Foreign-Owned 5.276 4.797 3.432 5.350 19.344 -0.241
(Dummy) (4.010) (3.856) (3.737) (4.033) (13.101) (2.820)
Foreign-Owned* -1.519 -1.365 -1.394 -1.514 -4.472 -0.324
∆Working Conditions (0.866)* (0.973) (0.733)* (0.864)* (1.894)** (0.631)
Total Effect of WC in 0.800 0.824 0.714 0.807 0.542 0.880
Foreign-Owned Firms (0.208)*** (0.223)*** (0.179)*** (0.213)*** (0.318)* (0.251)***
Constant -3.080 18.045 -1.485 -1,283.65 -19.768 10.304
(4.092) (22.630) (3.978) (3122.476) (15.466) (4.731)**
Observations 614 614 769 614 163 451
Firms 278 278 289 278 71 207
R2 0.10 0.16 0.07 0.10 0.20 0.10
* significant at 10%; ** significant at 5%; *** significant at 1%
Robust standard errors in parentheses. Regression results for Eq. 4 (column 1), Eq. 4 with monitor controls (2), Eq. 4
with unionization excluded (3), Eq. 4 with a continuous time variable (4), Eq. 4 for original firms only (5), and Eq. 4
excluding the original firms (6). Reported R2 values are R
2 within.
46
Table 6b: Regression Results - Disaggregated Working Conditions
Working
Conditions
WC*Foreign
Ownership
Total Effect of WC in
Foreign-Owned Factories
Paperwork 1.704 -1.413 0.291
(0.599)*** (0.608)** (0.124)**
FACB -1.396 1.623 0.227
(0.816)* (0.819)** (0.092)**
Internal Relations 0.414 -0.211 0.204
and Benefits (0.141)*** (0.174) (0.166)
OSH 0.086 0.024 0.11
(0.163) (0.098) (0.158) * significant at 10%; ** significant at 5%; *** significant at 1%
Robust standard errors in parentheses. Regression results for Eq. 3 with disaggregated working conditions variables.
Coefficients for controls not shown due to their similarity to those presented in Table 6a.
R2 Within: 0.14
47
Table 6c: Regression Results - Disaggregated Foreign Ownership
Country of
Ownership
Working
Conditions
WC*Country
of Ownership
Total Effect of WC
in Country's
Factories
West 2.312 -2.618 -0.306
(0.868)*** (0.952)*** (0.401)
China 2.312 -0.900 1.412
(0.868)*** (0.981) (0.467)***
Hong Kong 2.312 -1.671 0.641
(0.868)*** (0.929)* (0.384)*
Taiwan 2.312 -1.562 0.750
(0.868)*** (0.892)* (0.260)**
Korea 2.312 -2.040 0.273
(0.868)*** (1.028)** (0.560)
Malaysia 2.312 -2.143 0.169
(0.868)*** (0.952)** (0.429)
Singapore 2.312 -2.125 0.188
(0.868)*** (0.870)** (0.247)
Other Asia 2.312 -1.111 1.201
(0.868)*** (1.014) (0.575)**
Cambodia 2.314
(0.867)*** * significant at 10%; ** significant at 5%; *** significant at 1%
Robust standard errors in parentheses. Regression results for Eq. 3 with disaggregated foreign ownership variables.
Coefficients for controls not shown due to their similarity to those presented in Table 6a. R
2 Within: 0.11
P
Figure 2a
Working
Conditions
Wages
High Effort
Low Effort
M
N
48
Appendix A: Wage and Working Conditions Survey Questions Category Subcategory Q# IMS Question Original Question
FACB Shop Stewards 16300
Does management provide the
shop stewards with everything
required?
Has the employer
provided the shop
stewards with an office,
meeting room, working
materials and poster-
displaying site?
FACB Shop Stewards 16300
Does the employer allow
each shop steward two
hours per week to
perform his/her task
while maintaining
normal wages and
benefits?
FACB Shop Stewards 16400
Does management get
permission from the labour
ministry before dismissing
shop stewards?
Have any shop stewards
or candidates for shop
stewards been dismissed
from his/her
work/function (a) from
the employer (from
his/her work)?
FACB Shop Stewards 16400
Was this authorized by
the labour inspector?
FACB Shop Stewards 16500
Have the shop stewards been
consulted and given their
written opinion on
redundancy?
Have the shop stewards
been consulted and given
their written opinion on
redundancy? (Art. 284)
FACB Strikes 43100
Did management punish any
workers for participating in the
strike?
Did the employer impose
any sanctions on workers
participating in any
strike? (A. 333)
FACB Strikes 43400
Did management reinstate all
workers after the strike?
Were all workers
reinstated in their jobs at
the end of a strike?
FACB Strikes 43600
Did management pay the
striking workers' wages during
the strike?
If yes, did the employer
pay the wages of the
strikers for the duration
of the strike?
FACB Unions 30400
Can workers freely form and
join trade unions of their
choice?
Is there any indication
that workers are
prevented from forming
or joining a trade union
of their own choosing?
(C. 87/98)
FACB Unions 30500
Has management discriminated
against any worker because of
the worker's union membership
or union activities?
Is there any indication
that any worker has
suffered disadvantages
because of his/herunion
membership or union
activities: (see IMS 364
below)
FACB Unions 38900
Does management interfere
with workers or unions when
they draw up their constitutions
and rules, hold elections, or
organize their activities,
administration or finances?
Have workers/trade
unions been prevented
from: (see IMS 390
below)
49
FACB Unions 41500
Are workers free not to join the
union(s)?
Is there any indication
that workers are
threatened/coerced to
join a trade union?
FACB Unions 41600
Has management taken steps to
bring the union(s) under its
control?
Is there any indication
that the employer has
done things to bring the
union under the
employers' control or
domination?
FACB Unions 41700
Is any worker's job dependent
on the worker not joining a
union?
Is there any indication
that the workers' job is
dependent on not joining
a trade union?
Internal
Relations
and
Benefits Child Labor 28800
Have monitors verified the
employment of workers below
age 15?
Is there any indication of
employment of children
below the age of 12?
Internal
Relations
and
Benefits Child Labor 28800
Is there any indication of
employment of children
between the ages of 12
and 15?
Internal
Relations
and
Benefits Child Labor 29100
Does management keep a
register of workers who are
under age 18?
Does the employer keep
a register of employed
children below the age of
18?
Internal
Relations
and
Benefits Compensation for
Accidents/Illnesses 35300
(212. Does management
compensate workers correctly
for work-related accidents and
illnesses?) What types of
compensation owed to workers
has management failed to pay
correctly?
Do workers receive any
of the following forms of
compensation for work
related
accidents/illnesses?
Internal
Relations
and
Benefits Compensation for
Accidents/Illnesses 35302
costs for medication, treatment
and hospitalization
costs for medical care
and hospitalisation
Internal
Relations
and
Benefits Compensation for
Accidents/Illnesses 35303
annuity for permanently
disabled workers (20% or more
disabled)
an annuity for fatal
accidents or permanent
disability to the worker
or his/her beneficiaries
Internal
Relations
and
Benefits Compensation for
Accidents/Illnesses 35304
supplementary compensation
for permanently disabled
workers who require constant
care
supplementary
compensation for
permanently disabled
workers who require
constant care
Internal
Relations
and
Benefits Compensation for
Accidents/Illnesses 35305 funeral costs
costs for funerals and
survivors' pension
Internal
Relations
and
Benefits
Discipline/
Management
Misconduct 19300
Does management, including
line supervisors, treat workers
with respect?
Is there evidence of
indecent behaviour by
employers/managers?
Internal
Relations
and
Discipline/
Management
Misconduct 34701
(70. Does management make
any unauthorized deductions
from workers' wages?) What
What deductions are
made from wages? Fines
for misconduct/discipline
50
Benefits does management deduct?
disciplinary fines
(Art. 28)
Internal
Relations
and
Benefits
Discipline/
Management
Misconduct 44700
Are workers subject to
unwelcome conduct of a sexual
nature (physical contact,
spoken words, or conduct that
creates an intimidating or
humiliating work
environment)?
Is there evidence of
sexual harassment?
Internal
Relations
and
Benefits Discrimination 28100
Does management dismiss
pregnant workers or force them
to resign?
Is there evidence that
women have been fired
for becoming pregnant?
Internal
Relations
and
Benefits Discrimination 28400
Does management discriminate
against workers during hiring,
employment, or termination
based on their race, colour, sex,
religion, creed, ancestry, social
origin, or political opinion?
Is there any indication
that any worker has
suffered disadvantages
because of his/her race,
colour, sex, creed,
religion, political
opinion, birth, national
extraction, or social
origin: at the time of
recruitment? During
employment? At
termination of
employment?
Internal
Relations
and
Benefits Discrimination 44300
Does management dismiss
workers or change their
employment status or seniority
during maternity leave?
Have women been fired
during maternity leave or
at a date when the end
Internal
Relations
and
Benefits Disputes 17700
Has management implemented
the conciliation agreement?
If yes, was the agreement
implemented?
Internal
Relations
and
Benefits Disputes 18300
Did management implement
the arbitration award?
If yes, was the award
implemented
immediately?
Internal
Relations
and
Benefits Disputes 19000
Did management implement
conciliation agreements (if
any)?
If yes, was the agreement
implemented?
Internal
Relations
and
Benefits Forced Labor 28500
Is there any evidence of forced
(involuntary) labour?
Is there any evidence of
work being undertaken:
(see IMS 361 below)
Internal
Relations
and
Benefits Forced Labor 36100
In what form is forced labour is
occurring?
Is there any evidence of
work being undertaken:
Internal
Relations
and
Benefits Forced Labor 36105
labour as punishment for
holding views different from
mainstream political thought
as punishment for
holding views different
from management?
Internal
Relations
and Forced Labor 36106
labour as a means of labour
discipline
as a means of labour
discipline?
51
Benefits
Internal
Relations
and
Benefits
Holidays/
Annual Leave/
Special Leave 10400
Does management give
workers who have worked one
year or more any annual leave
at all (paid or unpaid) or any
annual leave compensation?
Is paid annual leave
given? (Art. 166)
Internal
Relations
and
Benefits
Holidays/
Annual Leave/
Special Leave 10500
Does management give
workers at least 18 days of paid
annual leave each year?
If yes, does this amount
to one and a half days per
month for continuous
service?
Internal
Relations
and
Benefits
Holidays/
Annual Leave/
Special Leave 11100
Do workers get 7 days of paid
special leave?
What is the maximum
amount of special leave
days a worker can take
per year?
Internal
Relations
and
Benefits
Holidays/
Annual Leave/
Special Leave 11600
Is the annual leave deducted
only from the same year during
which the worker took special
leave?
If the worker has taken
all his/her annual leave,
does the employer deduct
the special leave taken
from the workers' annual
leave for the next year?
Internal
Relations
and
Benefits Liaison Officer 16600
Has management appointed a
liaison officer?
Has an independent and
neutral liaison officer
been appointed/recruited
by the employer?
(SARACHOR NO.
21/SRC/MOSALVY,
1999)
Internal
Relations
and
Benefits Liaison Officer 16700
Did management consult with
workers before appointing the
liaison officer?
If yes, were workers'
consulted prior to the
appointment?
Internal
Relations
and
Benefits Liaison Officer 17000
Do workers have easy access to
the liaison officer?
If yes, do workers have
easy access to the liaison
officer?
Internal
Relations
and
Benefits Maternity Benefits 12700
Do women workers get at least
90 days of maternity leave?
Do women workers
receive maternity leave
of 90 days? (Art. 182)
Internal
Relations
and
Benefits Maternity Benefits 12800
Do women workers who have
worked for more than one year
get paid for maternity leave?
Have workers that
receive no wages during
maternity leave been
inservice for a period of
one uninterrupted year?
Internal
Relations
and
Benefits Maternity Benefits 13000
Can women do light work for
two months after returning
from maternity leave?
Do women do light work
for a period of two
months after their
maternity leave?
Internal
Relations
and
Benefits Maternity Benefits 13200
Does management give
workers one hour of paid time
off for breast-feeding?
Is time-off for
breastfeeding provided
for workers that have
given birth less than one
year ago?
Internal
Relations
and
Benefits Maternity Benefits 13600
Does management pay the
childcare costs of women
employees?
If there is no day care
centre for children older
than 18 months, does the
employer pay female
workers for the charges
52
for placing their children
a day care centre?
Internal
Relations
and
Benefits Maternity Benefits 34900
Do women receive the proper
pay/benefits for maternity
leave?
Do women receive the
proper pay/benefits for
maternity leave?
Internal
Relations
and
Benefits Overtime
Accommodation 10200
Does management provide
transportation or a place to
sleep for workers who finish
work between 22:00 and
05:00?
If yes, are workers that
work anywhere between
2200 and 0500 provided
with a place to sleep
when they finish?
Internal
Relations
and
Benefits Overtime
Accommodation 10200
If no, are these workers
provide with
transportation when they
finish?
Internal
Relations
and
Benefits Regular Hours/
Weekly Rest 7600
Are normal working hours
more than 8 hours per day, 6
days per week?
What are the normal
hours of work? (Art.137)
Internal
Relations
and
Benefits Regular Hours/
Weekly Rest 7700
Does management give
workers at least 24 consecutive
hours off per week?
Is there a weekly rest
break of at least 24
consecutive hours?
Internal
Relations
and
Benefits Regular Hours/
Weekly Rest 8601 voluntary
If yes, have workers
voluntarily agreed to do
so?
Internal
Relations
and
Benefits Regular Hours/
Weekly Rest 8602 exceptional
If yes, is this for
exceptional and urgent
jobs?
Internal
Relations
and
Benefits Regular Hours/
Weekly Rest 8603 limited to 2 hours per day
If yes, on average how
many hours per week?
Internal
Relations
and
Benefits Regular Hours/
Weekly Rest 9401
Work on holidays is not:
voluntary
(20a. Are workers aware
of their official holidays
as determined by
MOSALVY? 20b. If
yes, do they work on
these days?) If yes, have
they voluntarily agreed to
do so?
OSH Chemicals 23900
Are chemicals properly stored
in a separate area of the
workplace?
Are chemicals properly
stored?
OSH Chemicals 24100
Does the factory have
satisfactory exhaust ventilation
in areas where chemicals are
used?
Has exhaust ventilation
been installed in areas
where chemicals are in
use?
OSH Chemicals 24100
Could exhaust ventilation
be improved?
OSH Chemicals 24300
Does management train
workers who work with
chemical substances how to use
them safely?
Are workers exposed to
dangerous substances
trained in the handling of
these substances?
OSH Chemicals 24600 Do workers who need it use the Do workers who need it
53
protective clothing and
equipment that is provided?
actually use this
[protective]
clothing/equipment?
OSH
Emergency
Preparedness 21300
Are procedures in place to
handle emergencies (e.g., fire,
explosion, natural disaster)?
Are procedures in place
to handle emergencies
(such as fire, explosion,
natural disaster)?
OSH
Emergency
Preparedness 21400
Are managers, supervisors and
workers aware of these
procedures?
Are managers,
supervisors and workers
aware of these
procedures?
OSH
Emergency
Preparedness 21500
Does the factory hold regular
emergency drills?
If yes, are regular
emergency drills held?
OSH
Emergency
Preparedness 21789
Are all emergency exit doors
clearly marked?
Are emergency exits
clearly marked,
accessible and unlocked?
OSH
Emergency
Preparedness 21789
Are all emergency exit doors
unlocked during working
hours, including overtime?
Are emergency exits
clearly marked,
accessible and unlocked?
OSH
Emergency
Preparedness 21789
Are all emergency exit doors
accessible?
Are emergency exits
clearly marked,
accessible and unlocked?
OSH
Emergency
Preparedness 22100
Are there enough regularly
serviced fire extinguishers
within easy reach of workers?
Are fire extinguishers
within easy reach of
workers?
OSH
Emergency
Preparedness 22100
Are fire extinguishers
regularly serviced?
OSH Health/First Aid 13400
Does the factory have a
functioning and accessible
nursing room?
Is a nursing room
provided in or near the
enterprise (for those
enterprises employing
100 or more women, art
186)?
OSH Health/First Aid 22400
Are there enough properly
stocked first-aid boxes in the
workplace that are easily
accessible to workers?)
Is a properly stocked first
aid kit available?
OSH Health/First Aid 22500
Does management provide
periodic first aid training to
workers?
Is there a trained person
available to provide first
aid?
OSH Health/First Aid 22600
Does the factory have an
infirmary? (if factory has less
than 50 workers, tick N/A)
Does the enterprise (if
employing more than 50
workers) have a
permanent infirmary?
(A. 242)
OSH Health/First Aid 22700
Does the infirmary have
enough beds?
(Information was given
in the comment space for
question 74)
OSH Health/First Aid 22900
Does the infirmary have
enough medicine and medical
equipment?
Is the infirmary equipped
to provide emergency
care?
OSH Health/First Aid 23100
Do workers have to pay for
medicine or treatment provided
by the infirmary?
Is it [treatment by the
infirmary] subject to an
restrictions/fees?
OSH Machine Safety 25400
Are the machines well
maintained?
Are machines regularly
maintained?
OSH Machine Safety 25600
Are proper guards installed on
all dangerous moving parts of
machines and power
Are proper guards
attached to all dangerous
moving parts of
54
transmission equipment? (not
including needle guards)
machines and power
transmission equipment?
OSH Machine Safety 25700
Are electrical wires and
switches properly installed?
Are electrical wires and
switches safe and in good
condition?
OSH Machine Safety 25800
Are electrical wires and
switches well maintained?
Are electrical wires and
switches regularly
maintained?
OSH Machine Safety 26000
Are transformers or earth
leakage devices used when
there is a danger of shock?
Are transformers or earth
leakage devices used
when there is a danger
of shock?
OSH Machine Safety 26100
Are workers trained to use
machines and equipment
safely?
Are workers trained in
the proper/safe use of
machines and
equipment?
OSH
Temperature/
Ventilation/
Noise/Light 24800
Is the workplace free of
reflection and glare?
Is lighting free of
reflection and glare?
OSH
Temperature/
Ventilation/
Noise/Light 24900
Are light fittings in good
condition?
Are light fittings in good
condition?
OSH
Temperature/
Ventilation/
Noise/Light 25200
Is hearing protection provided
to all workers who need it?
Is hearing protection
provided to all workers
who need it?
OSH
Temperature/
Ventilation/
Noise/Light 26500
Are heat levels in the factory
acceptable?
What are the results of
the temperature
measurements taken
throughout the factory
premises?
OSH
Temperature/
Ventilation/
Noise/Light 26600
Does the factory have adequate
ventilation and air circulation?
Is adequate ventilation
provided to all workers
throughout the factory?
OSH
Temperature/
Ventilation/
Noise/Light 26800
Are dust levels in the factory
acceptable?
What are the results of
the dust measurements
taken throughout the
factory premises?
OSH Welfare Facilities 13789
Does management provide safe
drinking water?
Does the workplace have
an adequate supply of
safe drinking water?
OSH Welfare Facilities 13789
Does management provide
enough drinking water?
Does the workplace have
an adequate supply of
safe drinking water?
OSH Welfare Facilities 13789
Are there enough drinking
water stations?
Does the workplace have
an adequate supply of
safe drinking water?
OSH Welfare Facilities 14400
Does management
unreasonably restrict workers
from drinking water?
Are there any restrictions
on drinking water?
OSH Welfare Facilities 14500
Does the factory have the
number of toilets required?
Does the factory have the
number of toilets
required?
OSH Welfare Facilities 15000
Are the toilets cleaned
regularly?
Are toilet facilities
regularly cleaned?
OSH Welfare Facilities 15100
Are the toilets close to the
workplace?
Are toilet and washing
facilities close to the
work area?
OSH Welfare Facilities 15200
Is enough soap and water
available near the toilets?
Is soap and water
available for washing?
55
OSH Welfare Facilities 15300
Does management
unreasonably restrict workers
from using the toilets?
Are there any restrictions
on toilet use?
OSH
Workplace
Operations 26970 Is the workplace clean?
Are all work areas and
access paths kept tidy
and clean?
OSH
Workplace
Operations 26970 Is the workplace tidy?
Are all work areas and
access paths kept tidy
and clean?
OSH
Workplace
Operations 27100
Are access paths wide enough
to allow for two-way traffic?
Are access paths wide
enough to allow two-way
traffic?
OSH
Workplace
Operations 27200
Are access paths free of
obstruction?
Are all work areas and
access paths free of
obstruction and hazards?
OSH
Workplace
Operations 27300
Is the surface of transport
routes even and not slippery?
Is the surface of transport
routes even and not
slippery?
OSH
Workplace
Operations 27400
Can workers easily reach
switches, controls, tools and
materials?
Are switches, tools,
controls and materials
placed within easy reach
of workers?
OSH
Workplace
Operations 27500
Do workers have enough
equipment for carrying heavy
or bulky materials?
Are workers provided
with push-carts and other
wheeled devices for
carrying heavy or bulky
materials
OSH
Workplace
Operations 27600
Do workers who work sitting
down have adjustable chairs
with backrests?
Are seated workers
provided with chairs with
a sturdy backrest?
OSH
Workplace
Operations 28000
Do workers have to bend over
or raise their hands to work
because the work height is not
adequately adjusted?
Is work height adjusted
to the needs of individual
workers to avoid
bending postures or high
hand positions?
Paperwork Collective
Agreements 17400
If there is no collective
agreement, did the parties
inform the labour inspector
about the collective dispute(s),
so the dispute(s) could be
conciliated?
If yes, but there is no
collective agreement, did
the parties notify the
labour inspector for
conciliation?
Paperwork Collective
Agreements 19600
Is the collective agreement at
least as good for workers as the
Labour Law?
If yes, how do the
provisions compare with
the Labour Code?
Paperwork Collective
Agreements 19900
Has management registered the
collective agreement with the
labour ministry?
If yes, has it been
properly registered (Art.
4 Prakas 197/98)
Paperwork Collective
Agreements 20100
Has management posted the
collective agreement in the
workplace?
If yes, has the registered
CA been posted
throughout the
establishment?
Paperwork
Communication
with Labor
Ministry 8000
Has management obtained the
required authorizations from
the labour ministry? (For
rotating weekly rest days)
(Has weekly time off
ever been suspended?) If
yes, and in case of rest by
rotating staff, have the
necessary authorisations
been obtained?
Paperwork Communication
with Labor 8900
Does management get
permission from the Labour
if yes, has the employer
requested MOSALVY
56
Ministry Inspector before workers work
overtime?
for such overtime to be
taken?
Paperwork
Communication
with Labor
Ministry 16900
Has management notified the
labour ministry about the
appointment of the liaison
officer?
If yes, has MOSALVY
been notified of the
appointment?
Paperwork
Communication
with Labor
Ministry 20900
Does management regularly
provide a summary report of
work-related accidents and
illnesses to the relevant
authorities?
Does the enterprise
notify the relevant
authorities of work
related
accidents/illnesses? (Art.
1 Prakas 58/98)
Paperwork
Communication
with Labor
Ministry 20900
If yes, do they do so
within the required 24
hours of the
accident/illness?
Paperwork Contracts/Hiring 1600
Do workers have to pay
someone to get a job?
Is there any indication
that workers had to pay
someone to
Paperwork Contracts/Hiring 2200
Do the employment contracts
specify the terms and
conditions of employment?
If yes, does it stipulate
the terms of
employment?
Paperwork Contracts/Hiring 34706
(70. Does management make
any unauthorized deductions
from workers' wages?) What
does management deduct? the
cost of a bond or guarantee to
get or keep the worker's job
What deductions are
made from wages? Job
placement fee
Paperwork Informing
Workers 5100
Has management posted
minimum wage information in
the workplace?
Has the minimum wage
been posted in the
workplace and in
payment and recruitment
offices? (Art. 109)
Paperwork Informing
Workers 5300
Does management provide
clearly written pay slips to
workers?
Do workers get a record
of wages paid to them?
Paperwork Informing
Workers 5500
Do workers understand the
calculation of wages?
If yes, do they
understand the wage
calculations?
Paperwork Informing
Workers 8100
Does management keep an up-
to-date list showing each
worker's schedule for weekly
time off?
If yes, and in case of rest
by rotating staff, is a
special list indicating the
names of workers and
their time off being kept
and updated?
Paperwork Informing
Workers 10300
Does management post the list
of public holidays in the
factory?
Are workers aware of
their official holidays as
determined by
MOSALVY?
Paperwork Informing
Workers 16800
Did management inform
workers about the appointment
of the liaison officer?
If yes, has the
appointment been
announced to the
workers?
Paperwork Internal
Regulations 100
Does the factory have internal
regulations?
Does the enterprise have
internal regulations?
(Art. 23 and Notice 9/97)
Paperwork Internal
Regulations 200
Do the internal regulations
comply with the labour law?
If yes, do they comply
with the labour law?
Paperwork Internal 300 Were worker representatives If yes, were workers
57
Regulations consulted on the internal
regulations when they were
written or amended?
consulted on the internal
regulations?
Paperwork Internal
Regulations 400
Have the internal regulations
been posted in the workplace?
If yes, have internal
regulations been
communicated to
workers?
Paperwork Internal
Regulations 500
Are the internal regulations
legible? If yes, what language?
Paperwork Internal
Regulations 500
If yes, are they placed in
a proper and accessible
place (such as work
place, application room )
and kept clean and
legible?
Paperwork Internal
Regulations 600
Have the internal regulations
been approved by a labour
inspector?
If yes, have internal
regulations been signed
off by a labour inspector?
Paperwork
Regular
Hours/Weekly
Rest 9900
Does management get
permission from the Labour
Inspector before suspending
the weekly break?
(19a. Has weekly time
off ever been
suspended?) If yes, is the
required authorisation
obtained prior to
suspension?
Paperwork Safety
Documentation 20200
Does the factory have a written
health and safety policy?
Does the enterprise have
a written policy or
guidelines on OSH?
Paperwork Safety
Documentation 20400
Is the health and safety policy
written in Khmer?
If yes, Is the policy
written in Khmer?
Paperwork Safety
Documentation 20600
Do workers and supervisors
understand the health and
safety policy?
If yes, is the policy
known to all workers and
supervisors?
Paperwork Safety
Documentation 20700
Has management posted safety
and health information in
Khmer (e.g., posters and signs)
in the workplace?
Are safety
posters/notices
displayed?
Paperwork Safety
Documentation 20700
If yes, are they written in
Khmer?
Paperwork Safety
Documentation 20800
Does management keep a
record of work-related
accidents and illnesses?
Does the enterprise keep
a record of accidents?
Paperwork Safety
Documentation 20800
Does the enterprise keep
a record of work-related
illnesses?
Paperwork Safety
Documentation 23500
Does management keep an
inventory of all chemicals
stored at the workplace?
Is there an inventory kept
of all chemicals on the
work site?
Paperwork Safety
Documentation 23600
Does management have safety
data sheets for chemicals used
at the workplace?
Are safety data sheets
held for chemicals kept
on site?
Paperwork Safety
Documentation 23700
Do workers understand the
content of the safety data
sheets?
Are workers aware of
and understand the
content of such data
sheets?
Paperwork Safety
Documentation 35200
Has management failed to take
steps to ensure workers'
occupational health and safety?